Voice retrieval method, device and equipment for process operation instruction book and medium

Abstract

The invention relates to the technical field of voice recognition, and provides a voice retrieval method, device, equipment and medium for a process operation instruction, and the method comprises the steps: carrying out triple noise reduction processing on audio data of a production workshop to obtain audio noise reduction data, and creating a voice retrieval environment of the production workshop according to the audio noise reduction data; extracting a wake-up word of the process operation instruction according to the voice retrieval environment, and triggering a voice interaction service according to the wake-up word; extracting a user instruction corresponding to the process operation instruction through the voice service threshold time of the voice interaction service, and performing voice recognition on the user instruction to obtain a voice recognition instruction; performing full-text retrieval on the process operation instruction according to the voice recognition instruction, and calculating the matching conformity of a retrieved result and a user instruction; and displaying the retrieval result to the target user according to the matching conformity, and extracting a retrieval feedback result of the target user. According to the method and the device, the operation instruction retrieval efficiency can be improved.

Description

Technical Field

[0001] The present invention relates to the field of speech recognition technology, and in particular to a speech retrieval method, device, equipment and medium for process operation instructions. Background Art

[0002] An operation instruction is a standardized document in the field of manufacturing that implements process technology and guides production participants to perform correct operations. Production participants carry out operations in sequence according to the work step requirements of the operation instruction. However, in order to improve the efficiency of retrieving the operation instruction, voice guidance is required for the operation instruction to retrieve the required content.

[0003] The existing work instruction book detection technology is that when it is needed to be consulted, it is laborious to search through a large number of bulky work instructions, and it may not be possible to quickly find the desired chapter, which is very inconvenient to use. Especially when it cannot be found after several searches, there will be risks of blindly starting work or making reckless moves, resulting in a very slow speed of consulting the work instructions, thereby reducing the efficiency of searching the work instructions. Summary of the invention

[0004] In view of the above problems, embodiments of the present invention provide a method, device, equipment and medium for voice retrieval of process operation instructions.

[0005] In a first aspect, an embodiment of the present invention provides a method for voice retrieval of a process operation instruction manual, comprising:

[0006] Performing triple noise reduction processing on the pre-acquired audio data of the production workshop to obtain audio noise reduction data, and creating a voice retrieval environment of the production workshop based on the audio noise reduction data;

[0007] Extracting a preset wake-up word of the process operation instruction book according to the voice retrieval environment, and triggering a voice interaction service according to the wake-up word;

[0008] Extracting the user instruction corresponding to the process operation instruction book through the voice service threshold time of the voice interaction service, performing voice recognition on the user instruction, and obtaining a voice recognition instruction;

[0009] Performing a full-text search of the process operation instruction according to the voice recognition instruction, and calculating the matching degree between the search result and the user instruction;

[0010] The search results are displayed to the target user according to the matching degree, and the search feedback results of the target user are extracted.

[0011] According to an embodiment of the present invention, the step of performing triple noise reduction processing on the pre-acquired audio data of the production workshop to obtain audio noise reduction data includes:

[0012] Obtaining the location of the noise source of the production workshop, and optimizing the terminal layout of the production workshop according to the location of the noise source;

[0013] Screening terminal devices in the optimized terminal layout;

[0014] Perform hardware noise reduction processing on the audio data of the production workshop through the screened terminal equipment;

[0015] The audio data after the hardware noise reduction processing is subjected to noise reduction processing to obtain audio noise reduction data.

[0016] According to an embodiment of the present invention, extracting a preset wake-up word of a process operation instruction book according to the voice retrieval environment includes:

[0017] Under the conditions of the voice retrieval environment, extracting the voice wake-up instruction of the target user;

[0018] Matching the voice wake-up instruction with the preset wake-up word of the process operation instruction book;

[0019] When the wake-up word is matched successfully, the successfully matched wake-up word is used as the wake-up word of the process operation instruction.

[0020] According to an embodiment of the present invention, extracting the user instruction corresponding to the process operation instruction book through the voice service threshold time of the voice interaction service includes:

[0021] Obtaining the interval time between the voice wake-up instruction and the user instruction;

[0022] When the interval time is not within the voice service threshold time, converting the state of the voice interaction service to a preset standby state;

[0023] When the interval time is within the voice service threshold time, a user instruction corresponding to the process operation instruction is received through the voice interaction service.

[0024] According to an embodiment of the present invention, the full-text search of the process operation instruction according to the voice recognition instruction includes:

[0025] Converting the speech recognition instruction into a text form instruction;

[0026] Dividing the text form instructions into instructions to obtain instruction action content and instruction retrieval content;

[0027] Matching the instruction action content with the preset retrieval action content through a preset regular matching formula;

[0028] When the command action content is matched successfully, extract the search keywords in the command search content;

[0029] Perform a full-text search of the process operation instructions according to the search keywords.

[0030] According to an embodiment of the present invention, the calculating the matching degree between the search result after the search and the user instruction includes:

[0031] Extracting search keywords from search results after searching;

[0032] Counting the number of matches between the search keyword and the instruction keyword in the user instruction;

[0033] The matching degree between the search result after the search and the user instruction is calculated according to the matching number.

[0034] According to an embodiment of the present invention, the step of performing noise reduction processing on the audio data after the hardware noise reduction processing to obtain the audio noise reduction data includes:

[0035] Extracting a frequency spectrogram corresponding to a time domain signal in the audio data;

[0036] Calculate the noise threshold of each frequency band of the time domain signal according to the spectrum diagram;

[0037] Performing noise suppression on the noise data in the audio data by using a preset mask and the noise threshold;

[0038] The audio data after noise suppression is used as audio noise reduction data.

[0039] In a second aspect, an embodiment of the present invention provides a voice search device for a process operation instruction manual, characterized in that it includes:

[0040] A triple noise reduction processing module, used to perform triple noise reduction processing on the pre-acquired audio data of the production workshop to obtain audio noise reduction data, and create a voice retrieval environment of the production workshop according to the audio noise reduction data;

[0041] A voice interaction service triggering module, used to extract a preset wake-up word of the process operation instruction book according to the voice retrieval environment, and trigger the voice interaction service according to the wake-up word;

[0042] A user instruction recognition module, used to extract the user instruction corresponding to the process operation instruction book according to the voice service threshold time of the voice interaction service, perform voice recognition on the user instruction, and obtain a voice recognition instruction;

[0043] A full-text search module, used to perform a full-text search on the process operation instruction according to the voice recognition instruction, and calculate the matching degree between the search result after the search and the user instruction;

[0044] The retrieval feedback result extraction module is used to display the retrieval results to the target user according to the matching degree, and extract the retrieval feedback results of the target user.

[0045] In a third aspect, an embodiment of the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the voice retrieval method for process operation instructions described in the above aspect.

[0046] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for voice retrieval of process operation instructions described in the above aspects.

[0047] Compared with the prior art, the above technical solution of the present invention has the following beneficial effects:

[0048] The embodiments of the present invention analyze the noise sources in the production workshop, optimize the layout of the sound pickup equipment, reduce noise through software and hardware, improve the sound pickup effect, and take triple noise reduction measures to make it possible for workshops in noisy environments to use voice services; define wake-up words, and users can use the wake-up words to wake up the voice service, thereby realizing the digitization of work instructions and providing a basis for voice interaction support; collect user voice instructions, perform voice recognition, analyze instructions, and execute voice instructions, thereby realizing the application of voice interaction; and retrieve and match user needs, present results, and respond to users through voice broadcasts, thereby realizing full-text retrieval in all work instructions and improving retrieval efficiency. Therefore, the voice retrieval method, device, equipment, and medium for process work instructions proposed in the present invention can solve the problem of low efficiency when retrieving work instructions. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0050] Figure 1 A workflow diagram of a method for voice retrieval of a process operation instruction manual according to a first embodiment of the present invention is shown;

[0051] Figure 2aA schematic diagram showing the original sound waveform of a production workshop according to the first embodiment of the present invention;

[0052] Figure 2b A schematic diagram showing the sound waveform after hardware noise reduction in a production workshop according to the first embodiment of the present invention;

[0053] Figure 2c A schematic diagram showing the sound waveform after software noise reduction in a production workshop according to the first embodiment of the present invention;

[0054] Figure 3 A functional module diagram showing a voice search device for process operation instructions according to a third embodiment of the present invention;

[0055] Figure 4 A schematic diagram of the structure of an electronic device for implementing the voice retrieval method for process operation instructions according to the fourth embodiment of the present invention is shown. DETAILED DESCRIPTION

[0056] The present disclosure is further described below in conjunction with the embodiments shown in the accompanying drawings.

[0057] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0058] The present invention proposes a voice retrieval method for process operation instructions based on voice recognition technology. Based on the voice recognition theory and combined with the voice noise reduction method, a voice interaction service is built to achieve full-text retrieval in all operation instructions and improve retrieval efficiency. Compared with traditional methods, the voice retrieval technology for process operation instructions is more efficient and reduces human subjectivity, and has great potential and application prospects in the full-text retrieval application of operation instructions.

[0059] Embodiment 1

[0060] like Figure 1 As shown, the present invention proposes a voice retrieval method for process operation instructions, comprising the following steps:

[0061] S1. Perform triple noise reduction processing on the pre-acquired audio data of the production workshop to obtain audio noise reduction data, and create a voice retrieval environment of the production workshop according to the audio noise reduction data.

[0062] In the embodiment of the present invention, the audio noise reduction data refers to data after the noise in the audio signal is removed or reduced, so as to improve the clarity and intelligibility of the audio in the production workshop.

[0063] In the embodiment of the present invention, the triple noise reduction processing is performed on the pre-acquired audio data of the production workshop to obtain the audio noise reduction data, including:

[0064] Obtaining the location of the noise source of the production workshop, and optimizing the terminal layout of the production workshop according to the location of the noise source;

[0065] Screening terminal devices in the optimized terminal layout;

[0066] Perform hardware noise reduction processing on the audio data of the production workshop through the screened terminal equipment;

[0067] The audio data after the hardware noise reduction processing is subjected to noise reduction processing to obtain audio noise reduction data.

[0068] In detail, there is a certain amount of noise in the production workshop. In order to use the voice of the process operation instruction more accurately in a better environment, it is necessary to reduce the noise of the audio data generated in the production workshop. The noise in the production workshop is reduced through triple noise reduction processing. First, through noise source analysis, according to the location and sound intensity of the noise source, the terminal layout of the equipment and the production workshop is optimized, such as keeping the noise source away from the area where voice communication is required or setting a sound barrier between the noise source and the voice communication area. Secondly, through terminal selection, a pickup device with better noise reduction effect is used to perform hardware noise reduction, such as using a microphone with noise suppression function, that is, a directional microphone or a microphone with active noise reduction function, so as to reduce the impact of background noise. Finally, the audio data is denoised again through an algorithm, so as to ensure the pickup effect through the triple noise reduction processing of optimizing the terminal layout, using a pickup device with better noise reduction effect, and reducing the noise of the voice again through an algorithm.

[0069] Specifically, Figure 2a The figure shows the original sound waveform of the production workshop. The horizontal axis represents time and the vertical axis represents the amplitude of the sound. The horizontal axis shows how the sound changes over time, and the vertical axis shows the intensity or volume of the sound signal. Figure 2a The maximum sound amplitude has reached 25000, and the sound collection time has reached 120000. It can be seen that there is a certain amount of noise data in the audio data of the production workshop. Therefore, it is necessary to perform noise reduction on the audio data of the production workshop. First, perform hardware noise reduction on the audio data, such as Figure 2b The figure shows the sound waveform after hardware noise reduction in the production workshop. Figure 2b The maximum sound amplitude in the video has exceeded 10,000, and the sound collection time has reached 40,000. It can be seen that the audio data of the production workshop still has some noise data after hardware noise reduction. Therefore, it is necessary to perform noise reduction on the audio data again according to the noise reduction algorithm, such as Figure 2c The following is a schematic diagram of the sound waveform after software noise reduction in the production workshop. Figure 2cThe maximum sound amplitude after software noise reduction does not exceed 10000, and the time of sound collection is between 30000 and 40000. The corresponding sound amplitude is significantly lower than the sound amplitude after hardware noise reduction, thereby reducing the noise in the production workshop and obtaining audio noise reduction data after noise reduction processing.

[0070] Furthermore, after the audio data of the production workshop is subjected to noise reduction processing, the voice retrieval environment of the production workshop is regenerated under the condition of the audio data after noise reduction processing. The voice retrieval environment refers to the environment of the production workshop where the audio data after noise reduction processing is located, thereby improving the accuracy and response speed of voice recognition. Performing voice recognition in the voice retrieval environment after noise reduction processing can reduce noise interference and help to accurately convey instructions. Therefore, accurate voice retrieval of process operation instructions can be performed in the voice retrieval environment.

[0071] S2. Extracting a preset wake-up word of the process operation instruction manual according to the voice retrieval environment, and triggering a voice interaction service according to the wake-up word.

[0072] In an embodiment of the present invention, the wake-up word refers to a specific word or phrase in the wake-up process operation instructions. The user needs to activate or wake up the voice interaction service by saying this word or phrase. The function of the wake-up word is to switch the device from standby state to active state so that the voice interaction service can start processing subsequent voice commands or perform interactive processing.

[0073] In an embodiment of the present invention, the step of extracting a preset wake-up word of a process operation instruction book according to the voice retrieval environment includes:

[0074] Under the conditions of the voice retrieval environment, extracting the voice wake-up instruction of the target user;

[0075] Matching the voice wake-up instruction with the preset wake-up word of the process operation instruction book;

[0076] When the wake-up word is matched successfully, the successfully matched wake-up word is used as the wake-up word of the process operation instruction.

[0077] In detail, under the conditions of a voice retrieval environment, the voice input of the target user is monitored and the wake-up instructions in the user's voice are extracted. The voice wake-up instructions of the target user can be identified through voice recognition technology to ensure that different accents and speaking speeds can be recognized, and the extracted voice wake-up instructions are matched with the wake-up words in the preset process operation instructions. The wake-up words can be matched through regular matching, where the wake-up words in the preset process operation instructions are a wake-up name, such as instructions, etc. Once the wake-up word is matched successfully, the matched wake-up word will be used as the wake-up word of the process operation instructions, so that the voice interaction service can be triggered according to the wake-up word, and subsequent interactive operations of the process operation instructions can be performed based on the wake-up word.

[0078] Specifically, according to the wake-up word, the relevant voice interaction service will be triggered. In the scenario where employees use work instructions, under normal circumstances, the voice interaction service of the work instruction system is in standby mode, and only one wake-up monitoring service is retained. When the employee calls the predefined wake-up word within a certain distance, the wake-up monitoring service receives the information, and then gives a user response voice broadcast, and wakes up the voice interaction service at the same time, waiting for further voice commands from the user. The voice interaction service is a technology that uses speech recognition, natural language processing (NLP) and speech synthesis (TTS) to achieve interaction between people and devices or systems. The voice interaction service enables users to communicate with the system through voice commands without the need for traditional manual input.

[0079] Furthermore, after the voice interaction service is triggered according to the wake-up word, the voice interaction device is activated, so that the device is ready to receive further voice commands, and the device switches to the interactive mode, allowing the user to perform more operations and command input. Therefore, it is necessary to receive user interaction commands after waking up the device, so as to identify and respond to the user's intentions, perform corresponding operations or provide the required information, provide an efficient interaction method, and improve user convenience and operating experience.

[0080] S3. Extract the user instruction corresponding to the process operation instruction book through the voice service threshold time of the voice interaction service, perform voice recognition on the user instruction, and obtain a voice recognition instruction.

[0081] In an embodiment of the present invention, the user instruction refers to a request or command issued by the user through voice, text or other input methods, which is used to perform a specific operation or provide required information, such as looking up the production process operation instruction of a certain component, finding the document titled "P12345 Component Production Process Operation Instructions", viewing the processing technology of component XYZ, etc.

[0082] In the embodiment of the present invention, the extracting the user instruction corresponding to the process operation instruction book through the voice service threshold time of the voice interaction service includes:

[0083] Obtaining the interval time between the voice wake-up instruction and the user instruction;

[0084] When the interval time is not within the voice service threshold time, converting the state of the voice interaction service to a preset standby state;

[0085] When the interval time is within the voice service threshold time, a user instruction corresponding to the process operation instruction is received through the voice interaction service.

[0086] In detail, when an employee calls a predefined wake-up word within a certain distance, the wake-up monitoring service receives the information, and then gives a user response voice broadcast, and simultaneously wakes up the voice interaction service, waiting for the user's further voice instructions. If no voice instructions are received within the predetermined threshold time, the voice interaction service enters the standby state. Within the voice service threshold time, the user speaks further instructions, and then the user's instructions corresponding to the process operation instructions within the voice service threshold time are extracted. The voice service threshold time determines the maximum time to wait for the user's actual instructions after receiving the wake-up instruction, and the voice service threshold time can be customized according to user needs.

[0087] Exemplarily, when a user issues a voice wake-up command (such as "Hey, assistant"), the timestamp at this time is recorded. After the user issues an actual command (such as "check the processing technology of component XYZ"), the timestamp is recorded again. The interval time is obtained by calculating the time difference from the wake-up command to the user command. The calculated interval time is compared with the voice service threshold time. If the interval time is greater than the voice service threshold time, the system will enter a standby state, indicating that the voice interaction device will pause or reduce its response to subsequent commands until the next wake-up command appears; if the interval time is within the threshold time, the voice interaction device continues to process the user's commands and performs related operations. For example, if the user issues a "check the processing technology of component XYZ" command within the specified time, the voice interaction device will execute the operation of checking the processing technology of component XYZ according to the command, thereby improving the efficiency of voice interaction, reducing misoperation, and saving resources when no commands are received for a long time.

[0088] Furthermore, after extracting the user instruction within the voice service threshold time, the voice interaction service will perform voice recognition on the user instruction so as to be able to perform corresponding operations on the process operation instructions through the recognized voice instruction.

[0089] In the embodiment of the present invention, the voice recognition instruction refers to a user command or request converted into text through voice recognition technology. The voice recognition instruction is input by the user through voice, and the voice interaction service will process it to execute the corresponding voice instruction operation.

[0090] In the embodiment of the present invention, the step of performing voice recognition on the user instruction to obtain the voice recognition instruction includes:

[0091] Performing noise reduction processing on the user command, and performing silent cutting on the user command after the noise reduction processing;

[0092] Performing a frame operation on the user command after the silence is removed, and extracting acoustic features of the user command after the frame operation;

[0093] The acoustic features are subjected to speech recognition using a preset acoustic model to obtain a speech recognition instruction.

[0094] Specifically, the voice interaction service performs noise reduction processing on the user's command to reduce the interference of background noise and improve the clarity of the voice signal. The noise reduction processing includes but is not limited to noise suppression and echo cancellation. After the user's command is subjected to noise reduction processing, the voice signal contains some silent parts at the beginning and end. By removing these silent parts, the voice interaction service can more accurately process the core content of the voice data and divide the voice signal with the silent parts at the beginning and end removed into small time windows (or "frames"), usually with a length of 20-40 milliseconds. Each frame contains the features of the voice signal in the time window, and extracts acoustic features from each frame. Acoustic features refer to Mel frequencies. Cepstral coefficients (MFCCs), filter bank features, etc. Acoustic features are used to represent the audio characteristics of speech signals for further processing and recognition, and the extracted acoustic features are input into the acoustic model for recognition, and the acoustic model is usually a neural network based on deep learning, such as long short-term memory network (LSTM) or convolutional neural network (CNN), which is used to recognize and understand the speech content. The recognition result may be further processed by a language model or decoder to improve the recognition accuracy and generate the final text output to obtain the speech recognition instruction, and then the user's voice instruction is converted into a text instruction that the machine can understand, so as to perform the corresponding operation or provide the required information.

[0095] Furthermore, the recognized voice command is sent to the process operation instruction system retrieval module. After receiving the command, the process operation instruction system retrieval module performs a full-text search according to the user's command. After receiving the text command, the retrieval module will parse the content of the command and identify the user's specific request, such as searching for "welding process instructions". Then, the retrieval module will perform a full-text search in the process operation instruction database, so that it can quickly find and provide the information required by the user from a large number of documents, thereby improving work efficiency and accuracy.

[0096] S4. Perform a full-text search on the process operation instruction according to the voice recognition instruction, and calculate the matching degree between the search result and the user instruction.

[0097] In the embodiment of the present invention, full-text search is a kind of information search, which is used to search for documents containing specific keywords or phrases in the process operation instructions, with the purpose of quickly finding relevant search content from a large amount of text.

[0098] In the embodiment of the present invention, the full-text search of the process operation instruction according to the voice recognition instruction includes:

[0099] Converting the speech recognition instruction into a text form instruction;

[0100] Dividing the text form instructions into instructions to obtain instruction action content and instruction retrieval content;

[0101] Matching the instruction action content with the preset retrieval action content through a preset regular matching formula;

[0102] When the command action content is matched successfully, extract the search keywords in the command search content;

[0103] Perform a full-text search of the process operation instructions according to the search keywords.

[0104] In detail, the instructions obtained by voice recognition are converted into text form, that is, the voice signal is converted into text form, and the instructions in text form are divided to separate the commands from the content, that is, the instruction action content (that is, the specific command action or operation to be executed) and the instruction retrieval content (that is, the information to be searched) are extracted, and the instruction action content includes but is not limited to opening, locating, and logging in, and then the instruction action content is matched with the predefined retrieval action through a regular expression. For example, if the retrieval action includes "find", "search", "display", etc., the regular expression can be designed to match these verbs and their variants, such as \b(find|search|display|show|retrieve)\b, and the instruction action content is obtained and cleaned, and redundant spaces, punctuation marks, etc. are removed to ensure that the regular expression can be matched correctly. For example, "Please find the file" is processed as "Please find the file", so as to match the instruction action content using a regular expression. After matching, check whether the match is successful, and confirm that the matched action is one of the predefined retrieval actions. According to the matching result, the corresponding retrieval action is executed. For example, if "find" is matched, the file search function will be started to find relevant entries in the process operation instruction.

[0105] Specifically, when the instruction action content matches successfully, the search keyword is extracted from the instruction search content. For example, if the search content is to search for the vehicle body, then the search is the command action, and the vehicle body is the search keyword in the search content. The search keyword refers to the content included after the command action, which is used to search for relevant information in the process instruction book, and the extracted search keyword is used to perform a full-text search on the process operation instruction book to find information or documents related to the keyword. For example, if the search keyword is the vehicle body, a full-text search is performed in the process operation instruction data according to the content of the vehicle body to retrieve the content related to the vehicle body, including the retrieved title, chapter, abstract and the location where the search keyword appears, so that different search content corresponding to the vehicle body can be retrieved.

[0106] Furthermore, when full-text search is performed according to user instructions, the corresponding search results after the full-text search are obtained by fuzzy search. Therefore, many search results will be obtained, and the matching degree of the many search results will be calculated to query the most accurate search results, thereby realizing full-text search in all work instructions and improving search efficiency.

[0107] In the embodiment of the present invention, the matching degree is an indicator used to measure the relevance or matching degree between the search result and the user instruction, which indicates the proportion of the part of the search result related to the user instruction in the entire search result.

[0108] In the embodiment of the present invention, the calculating of the matching degree between the search result after the search and the user instruction includes:

[0109] Extracting search keywords from search results after searching;

[0110] Counting the number of matches between the search keyword and the instruction keyword in the user instruction;

[0111] The matching degree between the search result after the search and the user instruction is calculated according to the matching number.

[0112] In detail, keywords are extracted from user instructions and search results, and the frequency of occurrence of the extracted keywords in the search results is calculated. Based on the occurrence of the keywords, the degree of conformity between the search results and the user instructions is calculated. For example, the degree of match can be represented by the ratio of the intersection size of the keywords to the total number of keywords in the user instructions.

[0113] For example, if the user instruction is "search for report files of wheel components", and the search results are report files of wheel components, technical specifications of wheel components, installation and maintenance guides of wheel components, quality control inspection results of wheel components, etc., then the extracted keywords of the user instruction are {search, wheel components, report files}, the search results for the keywords corresponding to report files of wheel components are {wheel components, report files}, the search results for the keywords corresponding to technical specifications of wheel components are {wheel components, technical specifications}, the search results for the keywords corresponding to installation and maintenance guides of wheel components are {wheel components, installation, maintenance guide}, and the search results for quality control inspection results of wheel components are {wheel components, installation, maintenance guide}. The keywords corresponding to the search result are {wheel components, quality control, inspection results}, and the search result is that the matching keywords of the wheel component report file are {wheel components, report file}, so the number of matching keywords is 2, namely {wheel components, report file}, and the total number of keywords of the user command is 3, namely (search, wheel components, report file), and the degree of compliance = the number of matching keywords / the total number of keywords of the user command = 2 / 3, and the search result is that the matching keyword of the technical specifications of the wheel components is 1, and the matching degree of the search result of the technical specifications of the wheel components is 1 / 3, thereby calculating the matching degree corresponding to each search result in turn.

[0114] Furthermore, the search results are displayed according to the calculated matching degree corresponding to the search results, so as to provide users with more accurate search results based on the search results, and realize efficient reference and use of process operation instructions.

[0115] S5. Display the search results to the target user according to the matching degree, and extract the search feedback results of the target user.

[0116] In an embodiment of the present invention, the search results are displayed to the target user according to the matching degree, the search results are sorted in descending order according to the matching degree, the final result is presented to the user, and the search results are voice broadcast, thereby ensuring that the user can view the most relevant information first, and the voice broadcast can enable the user to obtain information even if he cannot read the screen, such as when performing other activities, such as "The wheel component report file found contains detailed technical specifications, installation guidelines and maintenance instructions, please check the report for complete information."

[0117] Furthermore, users can selectively consult the retrieved results according to specific needs, thereby achieving efficient consultation and use of the work instructions, lowering the usage threshold, and improving the level of use of the work instructions.

[0118] In the embodiment of the present invention, the search feedback result refers to the feedback information on the search result received by the user after using the process operation instruction search system (such as a search engine, database query, document search, etc.).

[0119] In the embodiment of the present invention, extracting the search feedback result of the target user includes:

[0120] Extracting the search requirements of the target user and matching the search requirements with the search results;

[0121] When the matched search results meet the search requirements, the search feedback results of the target user are determined as successful search;

[0122] When the matched search results do not meet the search requirements, the search feedback result of the target user is determined as a search failure.

[0123] In detail, key requirements are extracted from the user's query request. For example, the user may need to find a specific process component information, technical documentation, instructions for use, etc. The requirements are classified to more accurately match the search results. The result list obtained after the user's query is extracted from the process operation instruction retrieval system, including title, abstract, link, publication information, etc. The extracted search requirements are compared with the search results item by item to determine whether the search results meet the user's search requirements, such as whether the search results cover the topics that the user is concerned about, whether the information in the search results accurately and comprehensively meets the user's needs, and whether the search results are within the time range specified by the user.

[0124] Specifically, if the search results accurately meet the user's search needs, the search is recorded as successful, that is, the information in the search results directly corresponds to the user's needs; if the search results fail to meet the user's search needs, the search is recorded as failed. Common failure signs include that the search results are irrelevant to user needs or the search results lack key information or are inaccurate. The matching results and feedback are classified and recorded for subsequent analysis. The records include user needs, search results, matching status (success / failure), and possible reasons for failure. By extracting and analyzing the search feedback results of target users, the performance and accuracy of the process operation instruction retrieval system can be improved.

[0125] Embodiment 2

[0126] In order to better understand the present invention, a second embodiment is used below to further explain the situation in which the embodiment of the present invention performs noise reduction processing on audio data.

[0127] In the embodiment of the present invention, the audio noise reduction data is audio data obtained through triple noise reduction processing.

[0128] In the embodiment of the present invention, the step of performing noise reduction processing on the audio data after the hardware noise reduction processing to obtain the audio noise reduction data includes:

[0129] Extracting a frequency spectrogram corresponding to a time domain signal in the audio data;

[0130] Calculate the noise threshold of each frequency band of the time domain signal according to the spectrum diagram;

[0131] Performing noise suppression on the noise data in the audio data by using a preset mask and the noise threshold;

[0132] The audio data after noise suppression is used as audio noise reduction data.

[0133] In detail, the audio signal in the time domain is converted to the frequency domain, that is, the time domain signal is converted into the frequency domain representation through short-time Fourier transform (STFT), then the time domain signal is decomposed into different frequency components and represented as a two-dimensional image of time and frequency, namely, a spectrogram. The spectrogram is the frequency component representation of the time domain signal in different time windows, which can reveal the distribution of noise and useful signals in the signal. In the spectrogram, the signal amplitude of each frequency band usually includes the target signal and noise, and then according to the extracted spectrogram, the noise threshold of each frequency band is calculated.

[0134] Specifically, in the spectrogram, silent segments or low signal-to-noise ratio (SNR) segments without major signals are selected to analyze the amplitude of the noise. The statistical characteristics of the spectral amplitudes of these segments (such as the mean or standard deviation) can be calculated or the noise level can be estimated using a known noise model. For example, it can be assumed that the noise is white noise or Gaussian noise and analyzed based on this assumption. The noise threshold, once the noise threshold is determined, a mask can be created to reduce or eliminate the impact of noise in the spectrogram. The role of the mask is to apply a threshold to each frequency band in the spectrogram to suppress or remove frequency components below the threshold, that is, according to the estimated noise threshold, a mask is generated for each frequency band. The mask can be a binary matrix indicating which frequency bands need to be retained and which need to be suppressed. The mask is applied to the spectrogram to set the amplitude of the frequency band with noise below the threshold to 0, or reduce its amplitude to the threshold to achieve noise suppression, so that the audio data after noise suppression processing is used as the audio noise reduction data after noise reduction processing.

[0135] The embodiments of the present invention analyze the noise sources in the production workshop, optimize the layout of the sound pickup equipment, reduce noise through software and hardware, improve the sound pickup effect, and take triple noise reduction measures to make it possible for workshops in noisy environments to use voice services; define wake-up words, and users can use the wake-up words to wake up the voice service, thereby realizing the digitization of work instructions and providing a basis for voice interaction support; collect user voice instructions, perform voice recognition, analyze instructions, and execute voice instructions, thereby realizing the application of voice interaction; and retrieve and match user needs, present results, and respond to users through voice broadcasts, thereby realizing full-text retrieval in all work instructions and improving retrieval efficiency. Therefore, the voice retrieval method, device, equipment, and medium for process work instructions proposed in the present invention can solve the problem of low efficiency when retrieving work instructions.

[0136] Embodiment 3

[0137] like Figure 3 As shown, this embodiment also provides a functional module diagram of a voice retrieval device for process operation instructions.

[0138] The voice retrieval device 100 for the process operation instruction manual described in this embodiment can be installed in an electronic device. According to the functions implemented, the voice retrieval device 100 for the process operation instruction manual may include a triple noise reduction processing module 101, a voice interaction service triggering module 102, a user instruction recognition module 103, a full-text retrieval module 104, and a retrieval feedback result extraction module 105. The module described in the present invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0139] In this embodiment, the functions of each module / unit are as follows:

[0140] The triple noise reduction processing module 101 is used to perform triple noise reduction processing on the pre-acquired audio data of the production workshop to obtain audio noise reduction data, and create a voice retrieval environment of the production workshop according to the audio noise reduction data;

[0141] The voice interaction service triggering module 102 is used to extract a preset wake-up word of the process operation instruction book according to the voice retrieval environment, and trigger the voice interaction service according to the wake-up word;

[0142] The user instruction recognition module 103 is used to extract the user instruction corresponding to the process operation instruction book according to the voice service threshold time of the voice interaction service, perform voice recognition on the user instruction, and obtain a voice recognition instruction;

[0143] The full-text search module 104 is used to perform a full-text search on the process operation instruction according to the voice recognition instruction, and calculate the matching degree between the search result and the user instruction;

[0144] The search feedback result extraction module 105 is used to display the search results to the target user according to the matching degree, and extract the search feedback results of the target user.

[0145] In detail, each module described in the voice retrieval device 100 for process operation instructions described in the embodiment of the present invention adopts the same technical means as the voice retrieval method for process operation instructions described in Example 1 and Example 2 when used, and can produce the same technical effects, which will not be repeated here.

[0146] Embodiment 4

[0147] like Figure 4 As shown, this embodiment also provides a computer electronic device, which may include a processor 10, a memory 11, a communication bus 12 and a communication interface 13, and may also include a computer program stored in the memory 11 and executable on the processor 10, such as a voice retrieval program for process operation instructions.

[0148] In some embodiments, the processor 10 may be composed of an integrated circuit, for example, a single packaged integrated circuit, or a plurality of packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and combinations of various control chips. The processor 10 is the control core (ControlUnit) of the electronic device, and uses various interfaces and lines to connect various components of the entire electronic device, and executes or executes programs or modules stored in the memory 11 (for example, voice retrieval programs for executing process operation instructions, etc.), and calls data stored in the memory 11 to execute various functions of the electronic device and process data.

[0149] The memory 11 includes at least one type of readable storage medium, and the readable storage medium includes a flash memory, a mobile hard disk, a multimedia card, a card-type memory (for example, an SD or DX memory, etc.), a magnetic memory, a disk, an optical disk, etc. In some embodiments, the memory 11 may be an internal storage unit of an electronic device, such as a mobile hard disk of the electronic device. In other embodiments, the memory 11 may also be an external storage device of an electronic device, such as a plug-in mobile hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device. Further, the memory 11 may also include both an internal storage unit of the electronic device and an external storage device. The memory 11 may not only be used to store application software and various types of data installed in the electronic device, such as the code of the voice retrieval program of the process operation instruction book, but also be used to temporarily store data that has been output or is to be output.

[0150] The communication bus 12 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The bus is configured to realize connection and communication between the memory 11 and at least one processor 10, etc.

[0151] The communication interface 13 is used for communication between the above-mentioned electronic device and other devices, including a network interface and a user interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between the electronic device and other electronic devices. The user interface may be a display (Display), an input unit (such as a keyboard (Keyboard)), and optionally, the user interface may also be a standard wired interface, a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, and an OLED (Organic Light-Emitting Diode, organic light-emitting diode) touch device, etc. Among them, the display may also be appropriately referred to as a display screen or a display unit, which is used to display information processed in the electronic device and to display a visual user interface.

[0152] The figure only shows an electronic device with components. Those skilled in the art will understand that the structure shown in the figure does not constitute a limitation on the electronic device, and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0153] For example, although not shown, the electronic device may also include a power source (such as a battery) for supplying power to each component. Preferably, the power source may be logically connected to the at least one processor 10 through a power management device, so that the power management device can realize functions such as charging management, discharging management, and power consumption management. The power source may also include one or more DC or AC power sources, recharging devices, power failure detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The electronic device may also include a variety of sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be repeated here.

[0154] It should be understood that the embodiment is for illustration only and the scope of the patent application is not limited to this structure.

[0155] The voice retrieval program of the process operation instruction stored in the memory 11 in the electronic device is a combination of multiple instructions. When running in the processor 10, it can achieve:

[0156] Performing triple noise reduction processing on the pre-acquired audio data of the production workshop to obtain audio noise reduction data, and creating a voice retrieval environment of the production workshop based on the audio noise reduction data;

[0157] Extracting a preset wake-up word of the process operation instruction book according to the voice retrieval environment, and triggering a voice interaction service according to the wake-up word;

[0158] Extracting the user instruction corresponding to the process operation instruction book through the voice service threshold time of the voice interaction service, performing voice recognition on the user instruction, and obtaining a voice recognition instruction;

[0159] Performing a full-text search of the process operation instruction according to the voice recognition instruction, and calculating the matching degree between the search result and the user instruction;

[0160] The search results are displayed to the target user according to the matching degree, and the search feedback results of the target user are extracted.

[0161] Specifically, the specific implementation method of the processor 10 for the above instructions can refer to the description of the relevant steps in the corresponding embodiment of the accompanying drawings, which will not be repeated here.

[0162] Furthermore, if the module / unit integrated in the electronic device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. The computer-readable storage medium can be volatile or non-volatile. For example, the computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, and a read-only memory (ROM).

[0163] Embodiment 5

[0164] This embodiment provides a storage medium storing a computer program. When the computer program is executed by a processor, the steps of the voice retrieval method for process operation instructions as described above are implemented.

[0165] These program codes can also be loaded onto a computer or other programmable data processing device so that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions executed on the computer or other programmable device for implementing the process. Figure 1 The steps of a specified function in a process or multiple processes.

[0166] Storage media include permanent and non-permanent, removable and non-removable media, and can be implemented by any method or technology to store information. Information can be computer-readable instructions, data structures, modules of programs or other data. Examples of storage media can include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission medium that can be used to store information that can be accessed by a computing device.

[0167] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

[0168] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0169] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0170] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0171] Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is not limited only according to the above description, and it is intended that all changes within the meaning and scope of equivalent elements within the scope of protection are included in the present invention.

[0172] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0173] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the system claim can also be implemented by one unit or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

[0174] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A voice retrieval method for process operation instructions, characterized in that: The method comprises: Performing triple noise reduction processing on the pre-acquired audio data of the production workshop to obtain audio noise reduction data, and creating a voice retrieval environment of the production workshop based on the audio noise reduction data; Extracting a preset wake-up word of the process operation instruction book according to the voice retrieval environment, and triggering a voice interaction service according to the wake-up word; Extracting the user instruction corresponding to the process operation instruction book through the voice service threshold time of the voice interaction service, performing voice recognition on the user instruction, and obtaining a voice recognition instruction; Performing a full-text search of the process operation instruction according to the voice recognition instruction, and calculating the matching degree between the search result and the user instruction; The search results are displayed to the target user according to the matching degree, and the search feedback results of the target user are extracted.

2. The voice retrieval method for process operation instructions according to claim 1, characterized in that: The triple noise reduction processing is performed on the pre-acquired audio data of the production workshop to obtain audio noise reduction data, including: Obtaining the location of the noise source of the production workshop, and optimizing the terminal layout of the production workshop according to the location of the noise source; Screening terminal devices in the optimized terminal layout; Perform hardware noise reduction processing on the audio data of the production workshop through the screened terminal equipment; The audio data after the hardware noise reduction processing is subjected to noise reduction processing to obtain audio noise reduction data.

3. The voice retrieval method for process operation instructions according to claim 1, characterized in that: The extracting a preset wake-up word of the process operation instruction book according to the voice retrieval environment includes: Under the conditions of the voice retrieval environment, extracting the voice wake-up instruction of the target user; Matching the voice wake-up instruction with the preset wake-up word of the process operation instruction book; When the wake-up word is matched successfully, the successfully matched wake-up word is used as the wake-up word of the process operation instruction.

4. The voice retrieval method for process operation instructions according to claim 3, characterized in that: The extracting the user instruction corresponding to the process operation instruction book through the voice service threshold time of the voice interaction service includes: Obtaining the interval time between the voice wake-up instruction and the user instruction; When the interval time is not within the voice service threshold time, converting the state of the voice interaction service to a preset standby state; When the interval time is within the voice service threshold time, a user instruction corresponding to the process operation instruction is received through the voice interaction service.

5. The voice retrieval method for process operation instructions according to claim 1, characterized in that: The full-text search of the process operation instruction according to the voice recognition instruction includes: Converting the speech recognition instruction into a text form instruction; Dividing the text form instructions into instructions to obtain instruction action content and instruction retrieval content; Matching the instruction action content with the preset retrieval action content through a preset regular matching formula; When the command action content is matched successfully, extract the search keywords in the command search content; Perform a full-text search of the process operation instructions according to the search keywords.

6. The voice retrieval method for process operation instructions according to claim 1, characterized in that: The calculating the matching degree between the search result after the search and the user instruction includes: Extracting search keywords from search results after searching; Counting the number of matches between the search keyword and the instruction keyword in the user instruction; The matching degree between the search result after the search and the user instruction is calculated according to the matching number.

7. The voice retrieval method for process operation instructions according to claim 2, characterized in that: The step of performing noise reduction processing on the audio data after the hardware noise reduction processing to obtain audio noise reduction data includes: Extracting a frequency spectrogram corresponding to a time domain signal in the audio data; Calculate the noise threshold of each frequency band of the time domain signal according to the spectrum diagram; Performing noise suppression on the noise data in the audio data by using a preset mask and the noise threshold; The audio data after noise suppression is used as audio noise reduction data.

8. A voice search device for process operation instructions, characterized in that: The device comprises: A triple noise reduction processing module, used to perform triple noise reduction processing on the pre-acquired audio data of the production workshop to obtain audio noise reduction data, and create a voice retrieval environment of the production workshop according to the audio noise reduction data; A voice interaction service triggering module, used to extract a preset wake-up word of the process operation instruction book according to the voice retrieval environment, and trigger the voice interaction service according to the wake-up word; A user instruction recognition module is used to extract the user instruction corresponding to the process operation instruction book according to the voice service threshold time of the voice interaction service, perform voice recognition on the user instruction, and obtain a voice recognition instruction; A full-text search module, used to perform a full-text search on the process operation instruction according to the voice recognition instruction, and calculate the matching degree between the search result after the search and the user instruction; The retrieval feedback result extraction module is used to display the retrieval results to the target user according to the matching degree, and extract the retrieval feedback results of the target user.

9. A computer device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the voice retrieval method for process operation instructions as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the voice retrieval method for process operation instructions described in any one of claims 1 to 7 are implemented.

Images