A smart laboratory management method based on AI voice control

Abstract

The present invention provides a smart laboratory management method based on AI voice control, comprising the following steps: S1: collecting audio and identifying it as source voice information; S2: transmitting the source voice information to a control center; S3: the control center identifies the source voice information as an electronic signal; S4: transmitting the electronic signal to a communication system; by identifying the collected audio as source voice information, determining the destination time interval of the source voice information, processing the corresponding information message within the time range contained in the destination time interval, and obtaining the electronic signal, it is not necessary to process all the message information in the source voice information, thereby reducing the amount of message information that needs to be parsed, improving the parsing efficiency, and avoiding the problem of long parsing time.

Description

Technical Field

[0001] The present invention relates to the field of information management and control technology, and in particular to a smart laboratory management method based on AI voice control. Background Art

[0002] With the development of science and technology, artificial intelligence (AI) has been widely used in various fields. Among them, the most popular is voice technology, which is no longer limited to simple voice recognition and synthesis, but has developed into a new dimension that can greatly improve people's lives and work efficiency. Voice technology can now not only help people communicate more conveniently, but also control home smart devices such as smart speakers, smart home systems, etc. through voice.

[0003] At present, the process of using AI voice technology in experiments is usually as follows: the laboratory's AI voice recognition device, control center, and communication system are connected in sequence. When the laboratory staff makes a call, the AI ​​voice recognition device recognizes the acquired audio into voice information and transmits it to the control center. The control center parses the voice information into electronic signals and transmits the electronic signals to the communication system to make a real-time call.

[0004] In the prior art, when the control center parses voice information into electronic signals, it usually collects voice information for a fixed period of time and then performs corresponding analysis on all voice information in that period of time. The amount of voice information is not small and each analysis takes a long time, resulting in low efficiency in voice information analysis. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and propose an intelligent laboratory management method based on AI voice control.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] An AI voice-based laboratory control device includes an AI voice recognition device, a control center, and a communication system; the control center includes an identification module, an association module, and an analysis module;

[0008] The identification module is used to identify the target time interval according to the source voice information received by the analysis program;

[0009] The association module includes a pre-set association relationship, wherein the association relationship refers to a one-to-one correspondence between an identification code of an information message and a specific time range;

[0010] The analysis module is used to analyze the information message to obtain the electronic signal;

[0011] The present invention also proposes a smart laboratory management method based on AI voice control, comprising the following steps:

[0012] S1: collect audio and identify it as source voice information;

[0013] The AI ​​speech recognition device collects audio in the laboratory in real time through a microphone or other audio input device; pre-processes the collected audio, including removing noise, increasing sound clarity, etc.;

[0014] After preprocessing, features are extracted from the audio through signal processing and machine learning techniques to convert the original audio into a form that can be processed by computers; the features include the spectrum, tone, time-frequency characteristics, etc. of the sound;

[0015] Recognize the speech content in the audio by using a pre-trained deep learning model or statistical model, and convert the recognized speech content into text or commands, where the text or commands are the source speech information;

[0016] S2: Transmit the source voice information to the control center;

[0017] After the AI ​​voice recognition device recognizes the collected audio as source voice information, it transmits the source voice information to the control center;

[0018] The source voice information is the voice message transmitted to the control center in sequence according to the time sequence;

[0019] The control center is located in a cloud server or a local data center.

[0020] S3: The control center recognizes the source voice information as an electronic signal;

[0021] The control center receives the source voice information and analyzes the source voice information, including the following sub-steps:

[0022] S31: Determine the destination time interval of the source voice information;

[0023] The control center uses an analysis program to analyze the source voice information, and determines the time range contained in each source voice information according to the time sequence of the source voice being transmitted to the control center, wherein the time range refers to the time period from the start time to the end time;

[0024] The identification module determines the target time interval in the source voice information according to actual needs; the target time interval refers to a specific time period that needs to be processed or analyzed from the start time to the end time;

[0025] S32: Segment the source voice information and obtain information messages;

[0026] Divide the acquired source voice information into P information regions, and obtain the total information in each information region, wherein the total information includes the minimum value, the maximum value, the mean value and the variance;

[0027] Calculate the minimum value, maximum value, mean value and variance of the source voice information in each area;

[0028] The minimum value refers to the smallest value among the binary values ​​of the source voice information in each information area;

[0029] The highest value refers to the largest value among the binary values ​​of the source voice information in each information area;

[0030] The mean value refers to the mean value of the binary value of the source voice information in each information area;

[0031] The variance refers to the variance of the binary values ​​of the source voice information in each information area;

[0032] Encapsulating the total message in the information area and the source voice information contained therein into an information message, wherein one information area corresponds to one information message, that is, P information messages, and setting identification codes for the P information messages in sequence, wherein the identification codes are unique;

[0033] Preferably, the source voice information capacity in each information area is the same;

[0034] S33: Determine the destination information message according to the destination time interval;

[0035] The database of the association module contains pre-set associations, which refer to the one-to-one correspondence between the identification code of the information message and the specific time range;

[0036] According to each time period that needs specific processing included in the destination time interval determined by the identification module in step S31, the identification code of the information message corresponding to each time period is retrieved from the database;

[0037] The information message corresponding to the identification code of the retrieved information message is the destination information message;

[0038] S34: parsing all the total messages in the destination information message to obtain the electronic signal of the source voice information corresponding to the destination time interval;

[0039] The following sub-steps are included:

[0040] S341: Setting a variable group, and moving the time of the variable group until it is the same as the time of the information message;

[0041] A variable group is randomly set, wherein the variable group is a time period consisting of a previous moment and a later moment, wherein the previous moment and the later moment are time points, and the previous moment is earlier than the later moment;

[0042] Compare the previous time and the next time of the variable group with the start time and the end time of the information message respectively;

[0043] If the previous time is equal to the start time of the information message, then it does not move;

[0044] If the previous time is not equal to the start time of the information message, move the previous time toward the start time until the previous time is the same as the start time;

[0045] If the later time is equal to the termination time of the information message, it will not be moved; if the later time is not equal to the termination time of the information message, the later time will be moved towards the termination time until the later time is the same as the termination time;

[0046] S342: parsing the value of the information message corresponding to the variable group to obtain an electronic signal;

[0047] The parsing module parses the total message of the information message between the previous moment and the next moment, and obtains the value contained in the information message, that is, the electronic signal;

[0048] Furthermore, a threshold of information messages is artificially set;

[0049] Compare the total number of messages in the information message to be parsed with the threshold value, and if it is less than the threshold value, obtain the electronic signal according to the method of steps S331-S332;

[0050] If it is greater than the threshold, a variable group is randomly set, where the variable group is a time period consisting of a previous moment and a later moment, where the previous moment and the later moment are time points, and the previous moment is earlier than the later moment;

[0051] The previous time and the next time are moved respectively until they are the same as the start time and the end time of the information message respectively;

[0052] The variable group is divided into a plurality of variable subgroups, each variable subgroup includes a preceding sub-time and a following sub-time, and the information message included in the time period between each preceding sub-time and the following sub-time is extracted;

[0053] The parsing module sequentially parses the total information of the information message corresponding to each variable subgroup to obtain the value in the information message, namely the electronic signal; and integrates the electronic signals of all variable subgroups to obtain the electronic signal of the information message.

[0054] S4: transmits electronic signals to the communication system;

[0055] The control center transmits the electronic signal to the communication system, and the communication system makes calls in real time based on the electronic signal.

[0056] Compared with the prior art, the present invention has the following beneficial effects:

[0057] The intelligent laboratory management method based on AI voice control proposed in the present invention recognizes the collected audio as source voice information, determines the destination time interval of the source voice information, processes the corresponding information message within the time range contained in the destination time interval, and obtains an electronic signal. It does not need to process all the message information in the source voice information, reduces the amount of message information that needs to be parsed, improves the parsing efficiency, and avoids the problem of time-consuming parsing. DETAILED DESCRIPTION

[0058] In order to provide a further understanding of the purpose, structure, features, and functions of the present invention, the following detailed description is given in conjunction with the embodiments.

[0059] An AI voice-based laboratory control device includes an AI voice recognition device, a control center, and a communication system; the control center includes an identification module, an association module, and an analysis module;

[0060] The identification module is used to identify the target time interval according to the source voice information received by the analysis program;

[0061] The association module includes a pre-set association relationship, wherein the association relationship refers to a one-to-one correspondence between an identification code of an information message and a specific time range;

[0062] The analysis module is used to analyze the information message to obtain the electronic signal;

[0063] The present invention also proposes a smart laboratory management method based on AI voice control, comprising the following steps:

[0064] S1: collect audio and identify it as source voice information;

[0065] The AI ​​speech recognition device collects audio in the laboratory in real time through a microphone or other audio input device; pre-processes the collected audio, including removing noise, increasing sound clarity, etc.;

[0066] After preprocessing, features are extracted from the audio through signal processing and machine learning techniques to convert the original audio into a form that can be processed by computers; the features include the spectrum, tone, time-frequency characteristics, etc. of the sound;

[0067] Recognize the speech content in the audio by using a pre-trained deep learning model or statistical model, and convert the recognized speech content into text or commands, where the text or commands are the source speech information;

[0068] By collecting audio and processing it in real time, we can quickly respond to events or instructions occurring in the laboratory. The preprocessing step removes noise and increases the clarity of the sound, which helps to improve the accuracy of the subsequent speech recognition model. Removing noise and optimizing sound quality can reduce the misrecognition rate and ensure that voice commands or information can be correctly understood and processed.

[0069] S2: Transmit the source voice information to the control center;

[0070] After the AI ​​voice recognition device recognizes the collected audio as source voice information, it transmits the source voice information to the control center;

[0071] The source voice information is the voice message transmitted to the control center in sequence according to the time sequence;

[0072] The control center is located in a cloud server or a local data center.

[0073] S3: The control center recognizes the source voice information as an electronic signal;

[0074] The control center receives the source voice information and analyzes the source voice information, including the following sub-steps:

[0075] S31: Determine the destination time interval of the source voice information;

[0076] The control center uses an analysis program to analyze the source voice information, and determines the time range contained in each source voice information according to the time sequence of the source voice being transmitted to the control center, wherein the time range refers to the time period from the start time to the end time;

[0077] The identification module determines the target time interval in the source voice information according to actual needs; the target time interval refers to a specific time period that needs to be processed or analyzed from the start time to the end time;

[0078] S32: Segment the source voice information and obtain information messages;

[0079] Divide the acquired source voice information into P information regions, and obtain the total information in each information region, wherein the total information includes the minimum value, the maximum value, the mean value and the variance;

[0080] Calculate the minimum value, maximum value, mean value and variance of the source voice information in each area;

[0081] The minimum value refers to the smallest value among the binary values ​​of the source voice information in each information area;

[0082] The highest value refers to the largest value among the binary values ​​of the source voice information in each information area;

[0083] The mean value refers to the mean value of the binary value of the source voice information in each information area;

[0084] The variance refers to the variance of the binary values ​​of the source voice information in each information area;

[0085] Encapsulating the total message in the information area and the source voice information contained therein into an information message, wherein one information area corresponds to one information message, that is, P information messages, and setting identification codes for the P information messages in sequence, wherein the identification codes are unique;

[0086] Preferably, the source voice information capacity in each information area is the same;

[0087] S33: Determine the destination information message according to the destination time interval;

[0088] The database of the association module contains pre-set associations, which refer to the one-to-one correspondence between the identification code of the information message and the specific time range;

[0089] According to each time period that needs specific processing included in the destination time interval determined by the identification module in step S31, the identification code of the information message corresponding to each time period is retrieved from the database;

[0090] The information message corresponding to the identification code of the retrieved information message is the destination information message;

[0091] S34: parsing all the total messages in the destination information message to obtain the electronic signal of the source voice information corresponding to the destination time interval;

[0092] The following sub-steps are included:

[0093] S341: Setting a variable group, and moving the time of the variable group until it is the same as the time of the information message;

[0094] A variable group is randomly set, wherein the variable group is a time period consisting of a previous moment and a later moment, wherein the previous moment and the later moment are time points, and the previous moment is earlier than the later moment;

[0095] Compare the previous time and the next time of the variable group with the start time and the end time of the information message respectively;

[0096] If the previous time is equal to the start time of the information message, then it does not move;

[0097] If the previous time is not equal to the start time of the information message, move the previous time toward the start time until the previous time is the same as the start time;

[0098] If the later time is equal to the termination time of the information message, it will not be moved; if the later time is not equal to the termination time of the information message, the later time will be moved towards the termination time until the later time is the same as the termination time;

[0099] S342: parsing the value of the information message corresponding to the variable group to obtain an electronic signal;

[0100] The parsing module parses the total message of the information message between the previous moment and the next moment, and obtains the value contained in the information message, that is, the electronic signal;

[0101] Furthermore, a threshold of information messages is artificially set;

[0102] Compare the total number of messages in the information message to be parsed with the threshold value, and if it is less than the threshold value, obtain the electronic signal according to the method of steps S331-S332;

[0103] If it is greater than the threshold, a variable group is randomly set, where the variable group is a time period consisting of a previous moment and a later moment, where the previous moment and the later moment are time points, and the previous moment is earlier than the later moment;

[0104] The previous time and the next time are moved respectively until they are the same as the start time and the end time of the information message respectively;

[0105] The variable group is divided into a plurality of variable subgroups, each variable subgroup includes a preceding sub-time and a following sub-time, and the information message included in the time period between each preceding sub-time and the following sub-time is extracted;

[0106] The parsing module parses the total information of the information message corresponding to each variable subgroup in turn to obtain the value in the information message, that is, the electronic signal; and integrates the electronic signals of all variable subgroups to obtain the electronic signal of the information message.

[0107] By pre-setting a threshold and comparing the total number of messages in information messages with the threshold, different processing methods are adopted according to the comparison results, thereby optimizing the efficiency of data processing. By selecting different processing methods (single processing or split processing) according to the total number of messages in information messages, the system has the ability to adapt to different data volumes and needs, and can meet the data processing requirements in different scenarios.

[0108] By calculating the total information carried in the target information message, the value of the voice information corresponding to the target time interval is obtained, and the electronic signal is obtained, which can improve the efficiency of the analysis operation and the processing speed of the voice information.

[0109] S4: transmits electronic signals to the communication system;

[0110] The control center transmits the electronic signal to the communication system, and the communication system makes calls in real time based on the electronic signal.

[0111] Preferably, an information capacity threshold of the information region is set in advance, and the information capacity in each information region is compared with the information capacity threshold. If the information capacity in the information region is lower than the information capacity threshold, the variance is calculated using the following formula:

[0112] ;

[0113] If the information capacity in the information area is not less than the information capacity threshold, the variance is calculated using the following formula:

[0114] ;

[0115] in, is the variance, is the information capacity in the information area, It is the first The binary value of the source voice information, is the mean of the binary values ​​of the source speech information in the information area;

[0116] The information content in the information area is analyzed using different calculation methods. While ensuring the accuracy of the information, it can enhance the efficiency of source voice information processing and improve the efficiency of the analysis operation.

[0117] The present invention has been described by the above-mentioned relevant embodiments, however, the above-mentioned embodiments are only examples for implementing the present invention. It must be pointed out that the disclosed embodiments do not limit the scope of the present invention. On the contrary, changes and modifications made without departing from the spirit and scope of the present invention are all within the scope of patent protection of the present invention.

Claims

1. A smart laboratory management method based on AI voice control, characterized by: The following steps are involved: S1: collect audio and identify it as source voice information; The AI ​​speech recognition device collects audio in the laboratory in real time, pre-processes the collected audio, extracts features from the audio, and recognizes the audio content and converts it into source speech information. Specifically, the AI ​​speech recognition device collects audio in the laboratory in real time through a microphone or other audio input device; pre-processes the collected audio, and the pre-processing includes removing noise and increasing sound clarity; After preprocessing, features are extracted from the audio through signal processing and machine learning techniques to convert the original audio into a form that can be processed by computers; the features include the spectrum, tone, and time-frequency characteristics of the sound; Recognize the speech content in the audio by using a pre-trained deep learning model or statistical model, and convert the recognized speech content into text or commands, where the text or commands are the source speech information; S2: Transmit the source voice information to the control center; After the AI ​​voice recognition device recognizes the collected audio as source voice information, it transmits the source voice information to the control center; The source voice information is the voice message transmitted to the control center in sequence according to the time sequence; The control center is located in a cloud server or a local data center; S3: The control center recognizes the source voice information as an electronic signal; The control center receives the source voice information and analyzes the source voice information; The following sub-steps are included: S31: Determine the destination time interval of the source voice information; The control center uses an analysis program to analyze the source voice information, and determines the time range contained in each source voice information according to the time sequence of the source voice being transmitted to the control center, wherein the time range refers to the time period from the start time to the end time; The identification module determines the target time interval in the source voice information according to actual needs; the target time interval refers to a specific time period that needs to be processed or analyzed from the start time to the end time; S32: Segment the source voice information and obtain information messages; Divide the acquired source voice information into P information regions, and obtain the total information in each information region, wherein the total information includes the minimum value, the maximum value, the mean value and the variance; Calculate the minimum value, maximum value, mean value and variance of the source voice information in each area; The minimum value refers to the smallest value among the binary values ​​of the source voice information in each information area; The highest value refers to the largest value among the binary values ​​of the source voice information in each information area; The mean value refers to the mean value of the binary value of the source voice information in each information area; The variance refers to the variance of the binary values ​​of the source voice information in each information area; Encapsulating the total message in the information area and the source voice information contained therein into an information message, wherein one information area corresponds to one information message, that is, P information messages, and setting identification codes for the P information messages in sequence, wherein the identification codes are unique; S33: Determine the destination information message according to the destination time interval; The database of the association module contains pre-set associations, which refer to the one-to-one correspondence between the identification code of the information message and the specific time period; According to the specific time period corresponding to the destination time interval determined by the identification module in step S31, the identification code of the information message corresponding to the specific time period is retrieved from the database; The information message corresponding to the identification code of the retrieved information message is the destination information message; S34: parsing all the total messages in the destination information message to obtain the electronic signal of the source voice information corresponding to the destination time interval, including the following sub-steps: S341: Setting a variable group, and moving the time of the variable group until it is the same as the time of the information message; S342: parsing the value of the information message corresponding to the variable group to obtain an electronic signal; S4: transmits electronic signals to the communication system; The control center transmits the electronic signal to the communication system, and the communication system makes calls in real time based on the electronic signal.

2. The intelligent laboratory management method based on AI voice control according to claim 1, characterized in that: In step S341, a variable group is randomly set, where the variable group is a time period consisting of a previous moment and a later moment, where the previous moment and the later moment are time points, and the previous moment is earlier than the later moment; Compare the previous time and the next time of the variable group with the start time and the end time of the information message respectively; If the previous time is equal to the start time of the information message, then it does not move; If the previous time is not equal to the start time of the information message, move the previous time toward the start time until the previous time is the same as the start time; If the later time is equal to the termination time of the information message, it will not be moved; if the later time is not equal to the termination time of the information message, the later time will be moved toward the termination time until the later time is the same as the termination time.

3. The intelligent laboratory management method based on AI voice control according to claim 1, characterized in that: In step S342, the parsing module parses the total message of the information message between the previous moment and the next moment to obtain the value contained in the information message, that is, the electronic signal; Furthermore, a threshold of information messages is artificially set; Compare the total number of messages in the information message to be parsed with the threshold value, and if it is less than the threshold value, obtain the electronic signal according to the method of steps S341-S342; If it is greater than the threshold, a variable group is randomly set, where the variable group is a time period consisting of a previous moment and a later moment, where the previous moment and the later moment are time points, and the previous moment is earlier than the later moment; The previous time and the next time are moved respectively until they are the same as the start time and the end time of the information message respectively; The variable group is divided into a plurality of variable subgroups, each variable subgroup includes a preceding sub-time and a following sub-time, and the information message included in the time period between each preceding sub-time and the following sub-time is extracted; The parsing module parses the total information of the information message corresponding to each variable subgroup in turn to obtain the value in the information message, that is, the electronic signal; and integrates the electronic signals of all variable subgroups to obtain the electronic signal of the information message.

4. An AI voice-based laboratory control device applied to claims 1-3, characterized in that: It includes an AI speech recognition device, a control center, and a communication system; the control center includes a recognition module, an association module, and a parsing module; The identification module is used to identify the target time interval according to the source voice information received by the analysis program; The association module includes a pre-set association relationship, wherein the association relationship refers to a one-to-one correspondence between an identification code of an information message and a specific time range; The analysis module is used to analyze the information message to obtain the electronic signal.