Intelligent integrated management system based on voice conversation and application method thereof

Abstract

The invention discloses an intelligent integrated management system based on voice dialogue and an application method thereof, and belongs to the technical field of multimedia processing, the application method is applied to the intelligent integrated management system of voice dialogue, and the intelligent integrated management system comprises a voice acquisition module used for acquiring voice data; the voice extraction module is used for carrying out keyword extraction on the collected voice data in real time and matching keywords with wake-up words; the voice storage updating module is used for adding the current voice segment acquired in real time into the storage voice segment and deleting the pre-added voice segment which is equivalent to the current voice segment in length; the voice analysis module is used for analyzing the voice information of the front segment and the rear segment of the current wake-up word by taking the current wake-up word in the stored voice segment as a center, and judging whether an execution instruction exists or not; the control module and the voice storage module judge that the execution instruction exists, output a control instruction and send a wake-up instruction at the same time; the objective of the invention is to avoid misrecognition during voice operation.

Description

Technical Field

[0001] The invention belongs to the technical field of multimedia processing, and relates to an intelligent comprehensive management system based on voice dialogue and an application method thereof. Background Art

[0002] With the continuous development of voice technology, intelligent systems based on voice dialogue are increasingly used in various fields, bringing great convenience to people's lives and work. However, there are still many challenges in achieving efficient and accurate voice interaction.

[0003] In practical applications, existing voice dialogue systems often have problems with misrecognition, which may lead to the execution of incorrect operating instructions, causing trouble or even losses to users. For example, in smart home control, misunderstanding of user voice commands may lead to incorrect operation of home appliances; in intelligent customer service, misrecognition may lead to the inability to accurately answer user questions, reducing service quality.

[0004] At present, there are several solutions to the problem of misrecognition of speech. Some systems use more advanced speech acquisition equipment to improve the quality of input speech. The advantage is that it can reduce the interference of environmental noise, but the disadvantage is that the cost is high and the effect of speech acquisition in complex environments is still limited; some systems improve the recognition accuracy by optimizing the speech recognition algorithm. The advantage is that it can improve the recognition effect to a certain extent, but it will increase the consumption of computing resources, and when the user is talking, the device will start working after the misrecognition command, causing frequent startup of the device, affecting the user experience. Summary of the invention

[0005] The present invention provides an intelligent integrated management system based on voice dialogue and an application method thereof, which can avoid the problem of misrecognition when implementing voice operation.

[0006] In order to solve the above problems, the technical solution adopted by the invention is: An intelligent integrated management system based on voice dialogue, including the following functional modules: Voice collection module, used for collecting voice data; The voice extraction module extracts keywords from the collected voice data in real time and matches the keywords with the wake-up words; A voice storage update module is used to store voice segments of preset length, add the current voice segment collected in real time to the stored voice segment and delete the previously added voice segment with the same length as the current voice segment; The voice analysis module analyzes the voice information before and after the current wake-up word within a preset time period, with the current wake-up word in the stored voice segment as the center, to determine whether there is an execution instruction; The control module and the voice storage module determine that there is an execution instruction, then output the control instruction and send out a wake-up instruction at the same time.

[0007] An application method of an intelligent integrated management system based on voice dialogue comprises the following steps: S01 collects voice data using a voice collection module; S02 processes the collected voice data in real time through the voice extraction module, extracts keywords, and matches the keywords with the preset wake-up words; S03 uses the voice storage update module to store the voice segments of preset length, adds the newly collected current voice segments, and deletes the previously added voice segments of equal length; S04: When the keyword extracted by the voice extraction module matches the wake-up word, the voice analysis module analyzes the preceding and following voice information within a preset time period, centering on the current wake-up word in the stored voice segment, to determine whether there is an execution instruction; S05: If the voice storage module determines that there is an execution instruction, the control module outputs the control instruction and sends a wake-up instruction at the same time.

[0008] The principle of this solution is that the voice extraction module quickly extracts key information from the collected voice data, matches the extracted keywords with the wake-up words, and preliminarily screens out voice segments that may contain valid instructions; the voice storage and update module continuously saves voice segments of preset length, and through a reasonable update mechanism, ensures that the stored voice data is both timely and can efficiently use storage space, retaining necessary information for subsequent in-depth analysis. When the keyword and the wake-up word are successfully matched, the voice analysis module takes effect, and with the current wake-up word in the stored voice segment as the core, it performs a detailed analysis of the voice information before and after it to determine whether there is an exact execution instruction. Finally, the control module quickly outputs the control instruction and issues the wake-up instruction based on the judgment result of the voice storage module when it determines that there is an execution instruction, so as to achieve precise control of related equipment.

[0009] The beneficial effects of this program: The voice extraction module can extract keywords from the collected voice data in real time and match them with wake-up words, which greatly improves the system's efficiency in capturing key information and reduces unnecessary information processing, so that it can make preliminary judgments on user intentions more quickly; the voice storage update module can store voice segments of preset lengths. Through continuous updates, it not only ensures the timeliness of voice data, but also makes rational use of storage space, providing sufficient information support for subsequent in-depth analysis. The voice analysis module, based on the matching of keywords and wake-up words, further conducts in-depth analysis of relevant voice information to determine whether there are execution instructions, effectively avoiding misoperation and improving the accuracy and reliability of the system's execution of instructions.

[0010] Furthermore, when the device is in the awake state, the voice storage update module does not perform data update. When the device is awake and in use, the data update operation is avoided from occupying system resources, thereby ensuring that the device can respond to the current voice interaction task stably and smoothly without freezes or delays, avoiding data updates while the device is processing voice input and output, reducing the risk of data conflicts and errors caused by simultaneous read and write operations, and ensuring the accuracy and reliability of voice interaction.

[0011] Furthermore, the voice analysis module is also used to extract the voiceprint information of the current wake-up word. If the voiceprint information of the execution keyword extracted from the voice data matches the voiceprint information of the current wake-up word, an execution instruction is issued. The execution instruction is only issued when the voiceprint information of the execution keyword matches the voiceprint information of the current wake-up word. This can effectively prevent unauthorized personnel from using the same keyword to control the device, enhance the security and privacy protection of device operation, avoid device misoperation caused by multiple people saying similar keywords in the same environment, and improve the accuracy and reliability of the device's execution of instructions.

[0012] Furthermore, the voice storage update module dynamically adjusts the storage time, automatically adjusts the preset time according to the frequency of voice input, and flexibly adjusts the storage time according to the frequency of voice input, so as to more effectively utilize storage resources. When the voice input is frequent, the storage time is appropriately extended to ensure that there is sufficient voice data for analysis and processing; when the input frequency is low, the storage time is shortened to avoid unnecessary resource occupation; and when the voice input frequency is high, a longer storage time can provide more contextual information, which helps to more accurately understand and process the user's voice commands, thereby improving the system's response speed and processing efficiency.

[0013] Furthermore, the control module has a self-learning function, which can optimize the output strategy of control instructions according to the user's usage habits and historical operation data. With the learning of user habits and historical data, the output of control instructions is more in line with the user's actual expectations, reducing the occurrence of misjudgments and erroneous operations, and improving the accuracy of instruction execution. The optimized output strategy can respond to the user's voice commands more quickly, reduce unnecessary confirmations and repeated operations, thereby improving the work efficiency of the entire system.

[0014] Furthermore, the voice analysis module is equipped with an emergency recognition mechanism. When specific emergency keywords or voice features are detected, relevant instructions are processed and executed with priority, emergency instructions are processed first, avoiding delays caused by queuing for ordinary instructions in emergency situations, and improving the efficiency and effectiveness of emergency handling.

[0015] Furthermore, the voice storage update module uses blockchain technology to encrypt and verify the stored voice data. Through the verification mechanism, it can ensure the integrity of the stored voice data, prevent data loss or damage, and reduce the risk of single point failure. Even if problems occur in some nodes, the data can still be restored from other nodes.

[0016] Furthermore, in S01, after the voice acquisition module acquires the voice data, the acquired voice data is preprocessed in real time, including denoising, gain adjustment and endpoint detection. Denoising can remove background noise in the environment, making the voice clearer and improving the intelligibility and quality of the voice. Gain adjustment can balance the volume of the voice, avoid instability in subsequent processing due to excessive volume fluctuations, and ensure the stability of the voice signal. Endpoint detection can accurately determine the starting and ending points of the voice, eliminate invalid silent parts, reduce the amount of data for subsequent processing, improve computing efficiency, and reduce unnecessary consumption of system resources, so that limited computing and storage resources can be more effectively used for key voice processing tasks.

[0017] Furthermore, in S03, the voice storage and update module uses distributed ledger technology to record the storage and update operations of the voice segments. The distributed ledger technology ensures the immutability and consistency of the records, ensures the integrity and accuracy of the voice segment storage and update operation records, and avoids malicious tampering or erroneous recording of data. All operation records are transparent and visible to relevant parties, which is convenient for auditing and tracking, and can clearly understand the processing history of voice data. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a system module flow chart of the present invention; Figure 2 The present invention is a flow chart of the method. DETAILED DESCRIPTION

[0019] The embodiment is basically as shown in the attached Figure 1-2 Shown An intelligent integrated management system based on voice dialogue and an application method thereof are provided to avoid the problem of misrecognition during voice operation.

[0020] The basic solution proposed in this plan is: An intelligent integrated management system based on voice dialogue, including the following functional modules: Voice collection module, used for collecting voice data; The voice extraction module extracts keywords from the collected voice data in real time and matches the keywords with the wake-up words; A voice storage update module is used to store voice segments of preset length, add the current voice segment collected in real time to the stored voice segment and delete the previously added voice segment with the same length as the current voice segment; The voice analysis module analyzes the voice information before and after the current wake-up word within a preset time period, with the current wake-up word in the stored voice segment as the center, to determine whether there is an execution instruction; The control module and the voice storage module determine that there is an execution instruction, then output the control instruction and send out a wake-up instruction at the same time.

[0021] An application method of an intelligent integrated management system based on voice dialogue, characterized in that it comprises the following steps: S01 collects voice data using a voice collection module; S02 processes the collected voice data in real time through the voice extraction module, extracts keywords, and matches the keywords with the preset wake-up words; S03 uses the voice storage update module to store the voice segments of preset length, adds the newly collected current voice segments, and deletes the previously added voice segments of equal length; S04: When the keyword extracted by the voice extraction module matches the wake-up word, the voice analysis module analyzes the preceding and following voice information within a preset time period, centering on the current wake-up word in the stored voice segment, to determine whether there is an execution instruction; S05: If the voice storage module determines that there is an execution instruction, the control module outputs the control instruction and sends a wake-up instruction at the same time.

[0022] The voice extraction module quickly extracts key information from the collected voice data, matches the extracted keywords with the wake-up words, and preliminarily screens out voice segments that may contain valid instructions; the voice storage and update module continuously saves voice segments of preset length, and through a reasonable update mechanism, ensures that the stored voice data is both timely and can efficiently use storage space, retaining necessary information for subsequent in-depth analysis. When the keyword and the wake-up word are successfully matched, the voice analysis module takes effect, and with the current wake-up word in the stored voice segment as the core, it performs a detailed analysis of the voice information before and after it to determine whether there are exact execution instructions. Finally, the control module quickly outputs control instructions and issues wake-up instructions based on the judgment result of the voice storage module when it determines that there are execution instructions, thereby achieving precise control of related equipment.

[0023] The voice extraction module can extract keywords from the collected voice data in real time and match them with wake-up words, which greatly improves the system's efficiency in capturing key information and reduces unnecessary information processing, so that it can make preliminary judgments on user intentions more quickly; the voice storage update module can store voice segments of preset lengths. Through continuous updates, it not only ensures the timeliness of voice data, but also makes rational use of storage space, providing sufficient information support for subsequent in-depth analysis. The voice analysis module, based on the matching of keywords and wake-up words, further conducts in-depth analysis of relevant voice information to determine whether there are execution instructions, effectively avoiding misoperation and improving the accuracy and reliability of the system's execution of instructions.

[0024] When the device is in the awake state, the voice storage update module does not perform data update. When the device is awake and in use, the data update operation is avoided from occupying system resources, thereby ensuring that the device can respond to the current voice interaction task stably and smoothly without freezing or delay, avoiding data update while the device is processing voice input and output, reducing the risk of data conflicts and errors caused by simultaneous read and write operations, and ensuring the accuracy and reliability of voice interaction.

[0025] The voice analysis module is also used to extract the voiceprint information of the current wake-up word. If the voiceprint information of the execution keyword extracted from the voice data matches the voiceprint information of the current wake-up word, an execution instruction is issued. The execution instruction is only issued when the voiceprint information of the execution keyword matches the voiceprint information of the current wake-up word. This can effectively prevent unauthorized personnel from using the same keyword to control the device, enhance the security and privacy protection of device operation, avoid device misoperation caused by multiple people saying similar keywords in the same environment, and improve the accuracy and reliability of the device's execution of instructions.

[0026] The voice storage update module dynamically adjusts the storage time, automatically adjusts the preset time according to the frequency of voice input, and flexibly adjusts the storage time according to the frequency of voice input, so as to more effectively utilize storage resources. When the voice input is frequent, the storage time is appropriately extended to ensure that there is sufficient voice data for analysis and processing; when the input frequency is low, the storage time is shortened to avoid unnecessary resource occupation; and when the voice input frequency is high, a longer storage time can provide more contextual information, which helps to more accurately understand and process the user's voice commands, thereby improving the system's response speed and processing efficiency.

[0027] The control module has a self-learning function and can optimize the output strategy of control instructions according to the user's usage habits and historical operation data. With the learning of user habits and historical data, the output of control instructions is more in line with the user's actual expectations, reducing the occurrence of misjudgments and erroneous operations, and improving the accuracy of instruction execution. The optimized output strategy can respond to the user's voice commands more quickly, reduce unnecessary confirmations and repeated operations, thereby improving the work efficiency of the entire system.

[0028] The voice analysis module is equipped with an emergency recognition mechanism. When specific emergency keywords or voice features are detected, relevant instructions are processed and executed with priority, emergency instructions are processed first, and delays caused by queuing for ordinary instructions in emergency moments are avoided, thereby improving the efficiency and effectiveness of emergency handling.

[0029] The voice storage update module uses blockchain technology to encrypt and verify the stored voice data. Through the verification mechanism, it can ensure the integrity of the stored voice data, prevent data loss or damage, and reduce the risk of single point failure. Even if problems occur in some nodes, the data can still be restored from other nodes.

[0030] In the S01, after the voice acquisition module acquires the voice data, the acquired voice data is preprocessed in real time, including denoising, gain adjustment and endpoint detection. Denoising can remove background noise in the environment, make the voice clearer, and improve the intelligibility and quality of the voice. Gain adjustment can balance the volume of the voice, avoid instability in subsequent processing due to excessive volume fluctuations, and ensure the stability of the voice signal. Endpoint detection can accurately determine the starting point and end point of the voice, exclude invalid silent parts, reduce the amount of data for subsequent processing, improve computing efficiency, and reduce unnecessary consumption of system resources, so that limited computing and storage resources can be more effectively used for key voice processing tasks.

[0031] In S03, the voice storage and update module uses distributed ledger technology to record the storage and update operations of the voice segments. The distributed ledger technology ensures that the records cannot be tampered with and are consistent, and ensures the integrity and accuracy of the voice segment storage and update operation records, thereby avoiding malicious tampering or erroneous recording of data. All operation records are transparent and visible to relevant parties, which facilitates auditing and tracking, and enables a clear understanding of the processing history of voice data.

[0032] Specifically, the voice collection module uses a highly sensitive microphone device to collect the voice data emitted by the user in real time. During the collection process, the voice signal is preliminarily amplified and filtered to improve the quality and clarity of the voice.

[0033] Voice extraction module: Uses advanced voice recognition algorithms and machine learning models to perform real-time analysis on the collected voice data, quickly extract key information, such as keywords, and accurately match these keywords with preset wake-up words. By continuously optimizing the matching algorithm, the accuracy and speed of matching are improved.

[0034] Voice storage update module: A storage unit with sufficient capacity is used to store voice segments of preset duration (for example, 30 seconds). During the storage process, the current voice segment collected in real time is added to the stored voice segment, and the first added voice segment with the same length as the current voice segment is deleted according to the first-in-first-out principle. When the device is in the wake-up state, the module suspends the data update operation, and the module can automatically adjust the preset duration according to the frequency of voice input. For example, when the voice input is frequent, the storage duration is appropriately extended to 60 seconds; when the input frequency is low, the storage duration is shortened to 20 seconds. In addition, blockchain technology is used to encrypt and verify the stored voice data, and distributed ledger technology is used to record the storage and update operations of voice segments.

[0035] Voice analysis module: When the keyword extracted by the voice extraction module successfully matches the wake-up word, within a preset time period (e.g. 10 seconds), the voice information before and after the current wake-up word in the stored voice segment is analyzed in depth. The voiceprint information of the current wake-up word is extracted through voiceprint recognition technology, and the voiceprint information of the execution keyword extracted from the voice data is strictly matched with the voiceprint information of the current wake-up word. The module is also equipped with an emergency recognition mechanism. When specific emergency keywords or voice features are detected, the relevant instructions are processed and executed first.

[0036] Control module: It has a self-learning function and can optimize the output strategy of control instructions according to the user's usage habits and historical operation data. Once the voice storage module determines that there is an execution instruction, it quickly outputs the corresponding control instruction and sends a wake-up instruction to the relevant equipment to ensure that the equipment can respond to and execute the user's voice instruction in time.

[0037] Specific application steps: S01: Use the high-sensitivity microphone of the voice acquisition module to collect the user's voice data. After the acquisition is completed, the voice data is immediately pre-processed in real time, including using digital filtering technology for denoising to remove background noise in the environment, making the voice clearer and improving the intelligibility and quality of the voice; adjusting the gain through the automatic gain control circuit to balance the volume of the voice, avoid instability in subsequent processing due to excessive volume fluctuations, and ensure the stability of the voice signal; using an algorithm based on energy and zero-crossing rate for endpoint detection to accurately determine the starting and ending points of the voice, eliminate invalid silent parts, reduce the amount of data for subsequent processing, and improve computing efficiency. Reduce unnecessary consumption of system resources so that limited computing and storage resources can be more effectively used for key voice processing tasks.

[0038] S02: The pre-processed voice data is processed in real time through the voice extraction module. Advanced voice recognition algorithms and machine learning models are used to quickly extract keywords and accurately match keywords with preset wake-up words.

[0039] S03: Use the voice storage update module to store a preset duration, such as a 30-second voice segment. Add the newly collected current voice segment and delete the previously added voice segment of the same length. Ensure that the stored voice segment is always timely and effective, and provide sufficient information support for subsequent analysis.

[0040] S04: When the keywords extracted by the voice extraction module match the wake-up word, the voice analysis module will perform a detailed analysis of the previous and following voice information within a preset time period, such as 10 seconds, with the current wake-up word in the stored voice segment as the center. Voiceprint recognition technology and natural language processing technology are used to deeply understand the semantics and contextual information in the voice and accurately determine whether there is an execution instruction.

[0041] S05: If the voice storage module determines that there is an execution instruction, the control module outputs the corresponding control instruction and sends a wake-up instruction at the same time. The control instruction will be accurately transmitted to the relevant device to achieve precise control of the device.

[0042] During the operation of the entire system: Through the efficient work of the voice extraction module, voice clips that may contain valid instructions can be quickly screened out, unnecessary information processing can be reduced, and a preliminary judgment of the user's intention can be made quickly.

[0043] The voice storage update module ensures the timeliness of voice data and efficiently utilizes storage space through a reasonable update and duration adjustment mechanism, providing sufficient and accurate information for subsequent in-depth analysis.

[0044] Based on the matching of keywords and wake-up words, the voice analysis module further analyzes the relevant voice information in depth to accurately determine whether there are execution instructions, effectively avoid misoperation, and improve the accuracy and reliability of command execution.

[0045] The control module uses self-learning function to optimize the output strategy of control instructions according to user usage habits and historical operation data, thereby improving the work efficiency and user experience of the entire system.

[0046] In emergency situations, the emergency recognition mechanism of the voice analysis module can give priority to relevant instructions to ensure efficient and timely emergency handling.

[0047] The above are only embodiments of the present invention. Common knowledge such as the known specific structures and characteristics in the scheme is not described in detail here. Ordinary technicians in the relevant field are aware of all the common technical knowledge in the technical field to which the invention belongs before the application date or priority date, can obtain all the existing technologies in the field, and have the ability to apply conventional experimental means before that date. Ordinary technicians in the relevant field can improve and implement this scheme in combination with their own abilities under the enlightenment given by this application. Some typical known structures or known methods should not become obstacles for ordinary technicians in the relevant field to implement this application. It should be pointed out that for those skilled in the art, several deformations and improvements can be made without departing from the structure of the present invention, which should also be regarded as the scope of protection of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. An intelligent integrated management system based on voice dialogue, including the following functional modules: Voice collection module, used for collecting voice data; The voice extraction module extracts keywords from the collected voice data in real time and matches the keywords with the wake-up words; The voice storage update module is used to store the stored voice segments of preset duration, add the current voice segment collected in real time to the stored voice segment and delete the previously added voice segment with the same length as the current voice segment; when the device is in the awake state, the voice storage update module does not perform data update; the voice storage update module adopts a method of dynamically adjusting the storage duration, and automatically adjusts the preset duration according to the frequency of voice input; the voice storage update module adopts blockchain technology to encrypt and verify the stored voice data; The voice analysis module, if the keyword extracted by the voice extraction module matches the wake-up word, within a preset time period, the voice information before and after the current wake-up word in the stored voice segment is analyzed to determine whether there is an execution instruction; the voice analysis module is also used to extract the voiceprint information of the current wake-up word, and if the voiceprint information of the execution keyword extracted from the voice data matches the voiceprint information of the current wake-up word, an execution instruction is issued; the voice analysis module is equipped with an emergency recognition mechanism, and when specific emergency keywords or voice features are detected, the relevant instructions are processed and executed with priority; The control module and the voice storage module determine that there is an execution instruction, then output the control instruction and send out a wake-up instruction at the same time.

2. According to claim 1, an intelligent integrated management system based on voice dialogue is characterized in that: The voice storage update module uses blockchain technology to encrypt and verify the stored voice data.

3. An application method of an intelligent integrated management system based on voice dialogue, characterized in that: The following steps are involved: S01 uses a voice collection module to collect voice data; after the voice collection module collects the voice data, the collected voice data is preprocessed in real time, including denoising, gain adjustment and endpoint detection; S02 processes the collected voice data in real time through the voice extraction module, extracts keywords, and matches the keywords with the preset wake-up words; S03 uses the voice storage update module to store the voice segments of preset length, adds the newly collected current voice segments, and deletes the previously added voice segments of equal length; S04: When the keyword extracted by the voice extraction module matches the wake-up word, the voice analysis module analyzes the preceding and following voice information within a preset time period, centering on the current wake-up word in the stored voice segment, to determine whether there is an execution instruction; S05: If the voice storage module determines that there is an execution instruction, the control module outputs the control instruction and sends a wake-up instruction at the same time.

4. The application method of the intelligent integrated management system based on voice dialogue according to claim 3 is characterized in that: In S03, the voice storage update module uses distributed ledger technology to record the storage and update operations of the voice segments.

Images