Coal mine intelligent scheduling method and device based on voice recognition, equipment and medium

Abstract

The invention relates to the technical field of intelligent scheduling, and discloses a coal mine intelligent scheduling method and device based on voice recognition, equipment and a medium, and the method comprises the steps: collecting a voice scheduling audio, and carrying out the noise reduction of the voice scheduling audio based on wavelet transform, and obtaining a to-be-recognized signal; performing semantic recognition to determine a voice scheduling scheme and extracting a scheduling index; collecting a current task index and a demand index, determining a current task range index, and determining a schedulable index; comparing the schedulable index with the scheduling index to judge whether the voice scheduling scheme is adjusted or not; and when it is judged that the voice scheduling scheme is adjusted, establishing a feature set to adjust the voice scheduling scheme, and obtaining a final voice scheduling scheme. Through the intelligent and automatic scheduling adjustment method, the precision and the response speed of coal mine production scheduling are improved, manual intervention is reduced, and the working efficiency and the safety are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent scheduling, and more specifically, to a coal mine intelligent scheduling method, device, equipment and medium based on speech recognition. Background Art

[0002] With the gradual scale and complexity of coal mine production, the production scheduling of coal mines has become increasingly important. Traditional coal mine scheduling methods usually rely on manual operations and empirical judgments, which are not only inefficient but also easily affected by human factors, resulting in scheduling errors or low efficiency. The core task of coal mine scheduling is to reasonably arrange work tasks based on real-time production conditions, equipment operation conditions, and personnel scheduling requirements to ensure the continuity and safety of production.

[0003] Currently, there are some problems in the coal mine scheduling process. For example, the noise interference in the coal mine environment is severe, and existing speech recognition systems often have difficulty working effectively in complex noise environments. In addition, the matching degree between the speech scheduling scheme and the actual task requirements and the intelligence level of the adjustment strategy are still relatively low, making it difficult to meet the dynamic and complex coal mine production scheduling requirements.

[0004] Therefore, it is necessary to design a coal mine intelligent scheduling method, device, equipment and medium based on speech recognition to solve the problems existing in the current technology. Summary of the Invention

[0005] In view of this, the present invention proposes a coal mine intelligent scheduling method, device, equipment and medium based on speech recognition, aiming to solve the problems that the matching degree between the current speech scheduling scheme and the actual task requirements and the intelligence level of the adjustment strategy are still relatively low, making it difficult to meet the dynamic and complex coal mine production scheduling requirements.

[0006] On the one hand, the present invention proposes a coal mine intelligent scheduling method based on speech recognition, including:

[0007] Collecting speech scheduling audio, and performing noise reduction processing on the speech scheduling audio based on wavelet transform to obtain a signal to be recognized;

[0008] Performing semantic recognition on the signal to be recognized, comparing the recognition features with a scheduling scheme database to determine a speech scheduling scheme and extracting a scheduling index in the speech scheduling scheme;

[0009] Collecting a current task index and a demand index, determining a current task range index according to the current task index, and determining a schedulable index according to the demand index and the current task range index;

[0010] Comparing the schedulable index with the scheduling index, and judging whether to adjust the speech scheduling scheme according to the comparison result;

[0011] When it is determined to adjust the voice scheduling scheme, the demand index, the scheduling index, and the schedulable index are established as a feature set, and the feature set is compared with the historical adjustment scheme, and the voice scheduling scheme is adjusted according to the comparison result to obtain the final voice scheduling scheme.

[0012] Further, when performing noise reduction processing on the voice scheduling audio based on wavelet transform to obtain a signal to be recognized, it includes:

[0013] Performing discrete wavelet transform on the voice scheduling audio using Daubechies 4 wavelet, performing 4-layer wavelet decomposition on the voice scheduling audio to obtain an approximation coefficient and a detail coefficient:

[0014] Using the soft threshold method to perform threshold denoising on the detail coefficient, and setting the threshold to 2 times the standard deviation of the signal noise;

[0015] Performing inverse wavelet transform using the denoised coefficient to reconstruct the denoised signal to obtain the signal to be recognized.

[0016] Further, when performing semantic recognition on the signal to be recognized, comparing the recognition features with the scheduling scheme database, and determining the voice scheduling scheme and extracting the scheduling index in the voice scheduling scheme, it includes:

[0017] Converting the signal to be recognized into text based on semantic recognition;

[0018] Performing word segmentation on the recognized text to extract keywords, where the keywords include resource type, starting position, target position, and scheduling index;

[0019] Comparing the recognized semantic features with the scheduling scheme database through keyword similarity matching, selecting the scheme with the highest similarity as the voice scheduling scheme, and determining the scheduling index according to the voice scheduling scheme.

[0020] Further, when judging whether to adjust the voice scheduling scheme according to the comparison result, it includes:

[0021] Collecting the normal operation records corresponding to the demand index, and extracting the corresponding demand index range from the normal operation records;

[0022] Comparing the scheduling index with the demand index range, and generating a current task range index according to the numerical size relationship, where the current task range index includes a left boundary value and a right boundary value;

[0023] Obtaining the index difference between the current task index and the left boundary of the current task range index, and using the index difference as the schedulable index;

[0024] When the schedulable index is less than or equal to the scheduling index, it is determined to adjust the voice scheduling scheme;

[0025] When the schedulable index is greater than the scheduling index, it is determined not to adjust the voice scheduling scheme.

[0026] Further, when comparing the feature set with the historical adjustment scheme and adjusting the voice scheduling scheme according to the comparison result, it includes:

[0027] When there is data in the historical adjustment scheme whose similarity with the feature set is greater than the similarity threshold, the scheduling index in the voice scheduling scheme is adjusted according to the historical adjustment coefficient corresponding to the historical adjustment scheme; when there is no data in the historical adjustment scheme whose similarity with the feature set is greater than the similarity threshold, a similarity set is obtained according to the historical adjustment scheme, and the scheduling index in the voice scheduling scheme is adjusted according to the adjustment coefficient determined according to the similarity set.

[0028] Further, when adjusting the scheduling index in the voice scheduling scheme according to the historical adjustment coefficient corresponding to the historical adjustment scheme, it includes:

[0029] When the data in the historical adjustment scheme whose similarity with the feature set is greater than the similarity threshold is unique, the scheduling index in the voice scheduling scheme is adjusted with the historical adjustment coefficient corresponding to the historical adjustment scheme;

[0030] When the data in the historical adjustment scheme whose similarity with the feature set is greater than the similarity threshold is not unique, the scheduling index in the voice scheduling scheme is adjusted with the average value of the historical adjustment coefficients corresponding to the historical adjustment scheme;

[0031] Further, when obtaining a similarity set according to the historical adjustment scheme and adjusting the scheduling index in the voice scheduling scheme according to the adjustment coefficient determined according to the similarity set, it includes:

[0032] Data in the historical adjustment scheme whose similarity with the feature set is greater than a% and less than the similarity threshold is established as a similarity set;

[0033] The historical adjustment coefficient corresponding to the data with the maximum similarity in the similarity set is selected as the basic adjustment coefficient, and the basic adjustment coefficient is corrected according to the historical adjustment coefficients of the remaining data whose similarity is greater than the similarity threshold, and the scheduling index in the voice scheduling scheme is adjusted with the corrected adjustment coefficient;

[0034]

[0035] Among them, represents the adjusted coefficient after correction, represents the number of data with similarity greater than the similarity threshold among the remaining data in the similar set, represents the i-th historical adjusted coefficient corresponding to the data with similarity greater than the similarity threshold among the remaining data in the similar set, represents the basic adjusted coefficient.

[0036] Compared with the prior art, the beneficial effects of the present invention are as follows: By performing noise reduction processing on the voice dispatch audio through wavelet transform, the recognition accuracy and reliability of the voice recognition system in a complex noise environment are improved, ensuring that voice dispatch can proceed smoothly in a high-noise environment such as a coal mine. Through semantic recognition of the signal to be recognized and comparison with the dispatch plan database, the voice command is automatically converted into a specific dispatch plan, and the key dispatch indexes in the dispatch plan are extracted. Further introducing the dynamic calculation of the current task index and the demand index, by determining the comparison between the schedulable index and the dispatch index, it is automatically judged whether the dispatch plan needs to be adjusted. The adjustment mechanism not only enhances the flexibility and adaptability of the dispatch plan, but also can optimize the production task in real time. The comparison of historical adjustment plans further improves the intelligence level of dispatch decision-making, enabling the dispatch system to make more accurate dispatch decisions based on experience and real-time data in a changing production environment. Through the intelligent and automated dispatch adjustment method, the accuracy and response speed of coal mine production dispatch are improved, manual intervention is reduced, and work efficiency and safety are enhanced.

[0037] On the other hand, the present application also provides a coal mine intelligent dispatch device based on voice recognition, which is applied to the above-mentioned coal mine intelligent dispatch method based on voice recognition, and includes:

[0038] A collection module, configured to collect voice dispatch audio, and perform noise reduction processing on the voice dispatch audio based on wavelet transform to obtain a signal to be recognized;

[0039] An identification module, configured to perform semantic recognition on the signal to be recognized, and compare the recognition features with the dispatch plan database to determine the voice dispatch plan and extract the dispatch indexes in the voice dispatch plan;

[0040] A processing module, configured to collect the current task index and the demand index, determine the current task range index according to the current task index, and determine the schedulable index according to the demand index and the current task range index;

[0041] A judgment module, configured to compare the schedulable index with the dispatch index, and judge whether to adjust the voice dispatch plan according to the comparison result;

[0042] The determination module is further configured to, when it is determined that the voice scheduling scheme needs to be adjusted, establish the demand index, the scheduling index, and the schedulable index as a feature set, compare the feature set with the historical adjustment scheme, and adjust the voice scheduling scheme according to the comparison result to obtain the final voice scheduling scheme.

[0043] On the other hand, the present application also provides a coal mine intelligent scheduling device based on voice recognition, including:

[0044] One or more processors;

[0045] A storage device for storing one or more programs,

[0046] When the one or more programs are executed by the one or more processors, the one or more processors implement the above-mentioned coal mine intelligent scheduling method based on voice recognition.

[0047] On the other hand, the present application also provides a computer-readable storage medium, and when the program is executed by a processor, the above-mentioned coal mine intelligent scheduling method based on voice recognition is implemented.

[0048] It can be understood that the above-mentioned coal mine intelligent scheduling method, device, equipment, and medium based on voice recognition have the same beneficial effects, which will not be elaborated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] By reading the detailed description of the preferred embodiments below, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present invention. Moreover, throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0050] Figure 1 is a flowchart of the coal mine intelligent scheduling method based on voice recognition provided by an embodiment of the present invention;

[0051] Figure 2 is a structural block diagram of the coal mine intelligent scheduling device based on voice recognition provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0052] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully conveyed to those skilled in the art. It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the drawings and in combination with the embodiments.

[0053] In some embodiments of the present application, referring to Figure 1 as shown, a coal mine intelligent scheduling method based on speech recognition includes:

[0054] S100: Collect voice scheduling audio, and perform noise reduction processing on the voice scheduling audio based on wavelet transform to obtain a signal to be recognized.

[0055] S200: Perform semantic recognition on the signal to be recognized, compare the recognition features with the scheduling plan database, determine the voice scheduling plan, and extract the scheduling index in the voice scheduling plan.

[0056] S300: Collect the current task index and demand index, determine the current task scope index according to the current task index, and determine the schedulable index according to the demand index and the current task scope index.

[0057] S400: Compare the schedulable index with the scheduling index, and determine whether to adjust the voice scheduling plan according to the comparison result.

[0058] S500: When it is determined to adjust the voice scheduling plan, establish the demand index, scheduling index, and schedulable index as a feature set, compare the feature set with the historical adjustment plan, and adjust the voice scheduling plan according to the comparison result to obtain the final voice scheduling plan.

[0059] Specifically, in S100, the dispatching audio signal is collected through speech recognition technology. In a complex coal mine environment, there is a large amount of noise interference, and traditional speech recognition methods are difficult to effectively recognize speech content. Wavelet transform is used to denoise the speech signal. The background noise is removed, and clear speech information is extracted to provide high-quality signal input for subsequent semantic recognition. Wavelet transform analyzes in the time domain and frequency domain, especially suitable for processing non-stationary signals, and has excellent noise suppression ability. In S200, after the denoising of the speech signal is completed, semantic analysis is performed on the signal to be recognized. Natural language processing (NLP) technology is used to perform semantic recognition on the speech content, and key information such as the dispatching area and the type of dispatching resources is extracted. Then, by comparing with the historical data in the dispatching scheme database, the best voice dispatching scheme is determined, and the dispatching index (such as resource requirements) in the dispatching scheme is extracted. In S300, after the voice dispatching scheme is determined, the current task index and demand index are collected in real time to evaluate the actual number of operators and the actual required number for the current task. Based on the current task index, the current task scope index (such as the range of personnel requirements, etc.) is further calculated, and then the dispatchable index is determined according to the demand index and the task scope index. This dispatchable index reflects the dispatchable data under the current environment and conditions. In S400, the calculated dispatchable index is compared with the dispatch index in the determined dispatch scheme. Whether the current dispatch scheme needs to be adjusted is judged through the comparison result. If the current dispatchable index does not match the dispatch index, it is judged that the dispatch scheme needs to be optimized and adjusted to ensure the best execution of the production task. In S500, if it is judged that the dispatch scheme needs to be adjusted, the current demand index, dispatch index, and dispatchable index are combined into a feature set. This feature set will be compared with the historical adjustment scheme, and through historical data and adjustment experience, the most suitable adjustment scheme is obtained, and the current dispatch scheme is optimized based on this. A new dispatch scheme is generated to ensure that the production task can be completed efficiently and smoothly.

[0060] It is understandable that through wavelet transform noise reduction processing, the problem of noise interference in the coal mine environment is effectively solved, the high accuracy and stability of the speech recognition system are ensured, and the limitations of traditional speech recognition in complex environments are overcome. By analyzing the speech input through an algorithm, the scheduling scheme is automatically matched, and dynamic adjustment is carried out in combination with the task requirements and the real-time data of the current production environment. It reduces manual intervention and improves the scheduling efficiency and accuracy. Through the dynamic calculation and comparison of the schedulability index, task scope index and demand index, it responds in real time to changes occurring in production, such as personnel transfer, etc., and quickly adjusts the scheduling scheme to maintain the continuity and safety of production. By using the comparison of historical adjustment schemes, in a constantly changing production environment, combined with past experience and successful cases, a more accurate scheduling scheme is provided, avoiding the disadvantages brought by simply relying on manual experience or static rules. By optimizing the scheduling scheme, improving the decision-making speed and accuracy, the efficiency of coal mine production scheduling is improved, and the safety risk caused by scheduling errors or delays is reduced.

[0061] In some embodiments of the present application, when obtaining the signal to be recognized by performing noise reduction processing on the speech scheduling audio based on wavelet transform, it includes: performing discrete wavelet transform on the speech scheduling audio using Daubechies 4 wavelet, performing 4-layer wavelet decomposition on the speech scheduling audio to obtain approximation coefficients and detail coefficients: using the soft threshold method to perform threshold denoising on the detail coefficients, and setting the threshold to 2 times the standard deviation of the signal noise. Using the denoised coefficients to perform inverse wavelet transform to reconstruct the denoised signal and obtain the signal to be recognized.

[0062] Specifically, the Daubechies 4 wavelet (D4) is a Daubechies wavelet with 4 wavelet coefficients and exhibits good performance in speech signal denoising. The discrete wavelet transform (DWT) decomposes a signal into different frequency subbands, capable of dividing a speech signal into "approximation coefficients" (low-frequency part) and "detail coefficients" (high-frequency part). The low-frequency part contains the main information of the signal, while the high-frequency part usually contains noise and details. By performing 4-layer wavelet decomposition on the voice scheduling audio, the speech signal is decomposed into multi-level sub-signals. Each layer of decomposition further refines the signal, separating signal components in different frequency ranges. This multi-layer decomposition can effectively capture the details of the speech signal and separate high-frequency noise from the low-frequency useful signal. Each layer of decomposition produces a set of approximation coefficients and detail coefficients. During the wavelet decomposition process, the detail coefficients contain the noise components of the signal. To remove the noise, a soft threshold denoising method is adopted. The soft threshold method corrects the detail coefficients by setting a threshold. Specifically, the set threshold is "2 times the standard deviation of the signal noise", that is, the denoising intensity is dynamically set by statistically calculating the standard deviation of the noise. If the magnitude of a certain detail coefficient is less than the set threshold, it will be "compressed" to zero; if the magnitude is greater than the threshold, the detail coefficient is retained and its magnitude is adjusted. Compared with the hard threshold denoising method, the soft threshold method has a smoother effect, can reduce the risk of over-reducing the signal intensity, and thus better retains the details of the speech signal. After denoising, the inverse wavelet transform (IDWT) is performed using the denoised coefficients (including the approximation coefficients and the threshold-processed detail coefficients) to reconstruct them back into the time-domain signal. In this way, the denoised signal is the signal to be recognized, retaining the main features of the speech and removing most of the noise, providing high-quality input for the subsequent speech recognition process.

[0063] It can be understood that through the multi-layer decomposition of wavelet transform and the soft threshold denoising method for detail coefficients, the speech signal and noise can be effectively separated. The soft threshold method can not only remove noise but also avoid over-denoising, retaining more detailed features of the speech, thereby improving the accuracy of speech recognition. The dynamic threshold setting method based on the noise standard deviation makes the denoising process highly adaptable. With the change of different noise environments, the automatic adjustment of the threshold can ensure that the denoising effect is always optimal and is applicable to the noise interference in various coal mine environments. Compared with the hard threshold method, the soft threshold method can process the signal more smoothly, avoiding the loss of important speech information. Especially in a complex background noise environment, it helps to ensure the integrity of the speech signal and thus improve the quality of subsequent speech recognition. Through the precise denoising process of the signal, it is ensured that the speech recognition system can operate stably in a high-noise environment without being interfered by external noise.

[0064] In some embodiments of the present application, when performing semantic recognition on the signal to be recognized, comparing the recognition features with the scheduling scheme database, determining the voice scheduling scheme, and extracting the scheduling index in the voice scheduling scheme, it includes:

[0065] Converting the signal to be recognized into text based on semantic recognition.

[0066] Performing word segmentation on the recognized text, and extracting keywords, where the keywords include resource type, starting position, target position, and scheduling index.

[0067] Comparing the recognized semantic features with the scheduling scheme database through keyword similarity matching, selecting the scheme with the highest similarity as the voice scheduling scheme, and determining the scheduling index according to the voice scheduling scheme.

[0068] Specifically, through speech recognition technology, the noise-reduced speech signal is converted into text. Relying on existing speech recognition technology, the language information in the speech signal is extracted and converted into text form, providing basic data for subsequent semantic analysis. After converting the speech signal into text, word segmentation is then performed on the text. After word segmentation, the keywords in the voice command are extracted. These keywords include but are not limited to: resource type: for example, equipment type, work type, etc.; starting position: referring to the starting point or working area of the scheduling task; target position: referring to the end point or target position of the scheduling task; scheduling index: the quantity of resource requirements. After extracting the keywords, these keywords are compared with the historical scheduling schemes in the scheduling scheme database. In order to determine the most suitable scheduling scheme, a keyword similarity matching method is adopted. Similarity matching is usually performed by calculating the vector space representation of the keywords or based on text similarity metrics (such as cosine similarity, Jaccard similarity, etc.). In this way, the similarity between the input voice content and the historical scheduling schemes is evaluated, and thus the scheme with the highest similarity is selected. After selecting the most similar scheduling scheme, the final scheduling index will be determined according to the scheduling information extracted from this scheme. The scheduling index is a parameter that measures the resource requirements of the scheduling task.

[0069] It can be understood that by means of word segmentation and keyword extraction techniques, the core information in the voice command is captured, including resource type, task location, resource requirements, etc., avoiding the ambiguity of the voice command and improving the practicability and accuracy of speech recognition. By performing similarity matching with the scheduling scheme database, the most suitable scheduling scheme can be quickly selected from historical data, reducing the need for manual intervention and improving decision-making efficiency and accuracy. By extracting and analyzing the keywords in the voice command, the scheduling scheme can be flexibly adjusted according to the specific requirements of the task, enabling it to adapt to different production environments and task requirements. By automatically generating a scheduling scheme from the voice command and adjusting it according to real-time requirements, changes in production (such as equipment failures, personnel transfers, etc.) can be quickly responded to, ensuring the efficient execution of production tasks.

[0070] In some embodiments of the present application, when determining whether to adjust the voice scheduling scheme according to the comparison result, it includes: collecting the normal operation records corresponding to the demand index and extracting the corresponding demand index range from the normal operation records. Comparing the scheduling index with the demand index range and generating the current task range index according to the numerical magnitude relationship, where the current task range index includes a left boundary value and a right boundary value. Obtaining the index difference between the current task index and the left boundary of the current task range index and using the index difference as the schedulable index.

[0071] Specifically, when the schedulable index is less than or equal to the scheduling index, it is determined to adjust the voice scheduling scheme. When the schedulable index is greater than the scheduling index, it is determined not to adjust the voice scheduling scheme.

[0072] It can be understood that by collecting and analyzing the demand index and normal operation records in real time, it automatically adapts to the changing requirements of different tasks and determines whether to adjust the scheduling scheme. By comparing the task requirements and resource scheduling capabilities, it ensures that resources are reasonably allocated to the greatest extent during the task scheduling process.

[0073] In some embodiments of the present application, when comparing the feature set with the historical adjustment scheme and adjusting the voice scheduling scheme according to the comparison result, it includes: when there is data in the historical adjustment scheme whose similarity to the feature set is greater than the similarity threshold, adjusting the scheduling index in the voice scheduling scheme according to the historical adjustment coefficient corresponding to the historical adjustment scheme. When there is no data in the historical adjustment scheme whose similarity to the feature set is greater than the similarity threshold, obtaining a similarity set according to the historical adjustment scheme and determining an adjustment coefficient according to the similarity set to adjust the scheduling index in the voice scheduling scheme.

[0074] In some embodiments of the present application, when adjusting the scheduling index in the voice scheduling scheme according to the historical adjustment coefficient corresponding to the historical adjustment scheme, it includes: when the data in the historical adjustment scheme with a similarity to the feature set greater than the similarity threshold is unique, adjusting the scheduling index in the voice scheduling scheme with the historical adjustment coefficient corresponding to the historical adjustment scheme. When the data in the historical adjustment scheme with a similarity to the feature set greater than the similarity threshold is not unique, adjusting the scheduling index in the voice scheduling scheme with the average value of the historical adjustment coefficients corresponding to the historical adjustment scheme.

[0075] In some embodiments of the present application, when obtaining a similarity set according to the historical adjustment scheme and determining an adjustment coefficient to adjust the scheduling index in the voice scheduling scheme, it includes: establishing a similarity set with the data in the historical adjustment scheme having a similarity to the feature set greater than a% and less than the similarity threshold.

[0076] Specifically, select the historical adjustment coefficient corresponding to the data with the maximum similarity in the similarity set as the basic adjustment coefficient, and correct the basic adjustment coefficient according to the historical adjustment coefficients in the remaining data with a similarity greater than the similarity threshold, and adjust the scheduling index in the voice scheduling scheme with the corrected adjustment coefficient.

[0077]

[0078] Among them, represents the corrected adjustment coefficient, represents the number of the remaining data in the similarity set with a similarity greater than the similarity threshold, represents the i-th historical adjustment coefficient corresponding to the remaining data in the similarity set with a similarity greater than the similarity threshold, represents the basic adjustment coefficient.

[0079] Specifically, after determining the adjustment coefficient to adjust the scheduling index in the voice scheduling scheme, multiply the adjustment coefficient by the scheduling index to obtain the final scheduling index, and combine the final scheduling index with the previously extracted scheduling scheme to obtain the final voice scheduling scheme. When using similarity combination to determine the corrected adjustment coefficient, the corrected adjustment coefficient and the corresponding feature set will be stored in the historical adjustment scheme to enrich the historical samples.

[0080] It is understandable that by using the empirical data in the historical adjustment plan to provide a highly targeted adjustment plan for the current task, the scheduling can be made more in line with the actual production requirements, reducing scheduling errors or resource waste. By selecting the data with the highest similarity in the similarity set as the basic adjustment coefficient and then weighted-correcting the historical adjustment coefficient, the influence of various historical data can be better balanced, avoiding the excessive bias of a single historical plan, thereby improving the accuracy and robustness of scheduling adjustment. By automatically selecting similar data and correcting the adjustment coefficient, automated scheduling optimization is achieved, without manual intervention, improving the intelligent level of production scheduling and reducing human decision-making errors. The correction of the similarity set and the historical adjustment coefficient can dynamically adjust the scheduling plan according to real-time data. No matter how the production environment changes, a flexible scheduling optimization plan can be provided, enhancing the adaptability of the production system. By finely adjusting the scheduling plan, resources can be better allocated and task conflicts reduced, thereby improving the efficiency of coal mine production, reducing downtime, and ensuring that tasks are completed on time.

[0081] In the above embodiments, the voice scheduling audio is denoised through wavelet transform, improving the recognition accuracy and reliability of the voice recognition system in a complex noise environment, ensuring that voice scheduling can proceed smoothly in the high-noise environment of a coal mine. Through semantic recognition of the signal to be recognized and comparison with the scheduling plan database, the voice command is automatically converted into a specific scheduling plan, and the key scheduling indices in the scheduling plan are extracted. Further introducing the dynamic calculation of the current task index and the demand index, by determining the comparison between the schedulable index and the scheduling index, it is automatically judged whether the scheduling plan needs to be adjusted. The adjustment mechanism not only enhances the flexibility and adaptability of the scheduling plan but also can optimize production tasks in real time. The comparison of the historical adjustment plan further improves the intelligent level of scheduling decision-making, enabling the scheduling system to make more accurate scheduling decisions based on experience and real-time data in a changing production environment. Through intelligent and automated scheduling adjustment methods, the accuracy and response speed of coal mine production scheduling are improved, manual intervention is reduced, and work efficiency and safety are increased.

[0082] In another preferred manner based on the above embodiments, referring to Figure 2 as shown, this embodiment provides a coal mine intelligent scheduling device based on voice recognition, which is applied to the above-mentioned coal mine intelligent scheduling method based on voice recognition and includes:

[0083] A collection module, configured to collect voice scheduling audio and obtain a signal to be recognized by performing denoising processing on the voice scheduling audio based on wavelet transform;

[0084] An identification module, configured to perform semantic recognition on the signal to be recognized, compare the recognition features with the scheduling plan database, determine the voice scheduling plan, and extract the scheduling indices in the voice scheduling plan;

[0085] A processing module, configured to collect the current task index and the demand index, determine the current task scope index according to the current task index, and determine the schedulable index according to the demand index and the current task scope index;

[0086] A judgment module, configured to compare the schedulable index with the scheduling index, and judge whether to adjust the voice scheduling scheme according to the comparison result;

[0087] The judgment module is further configured to, when it is determined to adjust the voice scheduling scheme, establish the demand index, the scheduling index, and the schedulable index as a feature set, compare the feature set with the historical adjustment scheme, and adjust the voice scheduling scheme according to the comparison result to obtain the final voice scheduling scheme.

[0088] On the other hand, the present application further provides a coal mine intelligent scheduling device based on voice recognition, including:

[0089] One or more processors;

[0090] A storage device, configured to store one or more programs,

[0091] When the one or more programs are executed by the one or more processors, the one or more processors implement the above-mentioned coal mine intelligent scheduling method based on voice recognition.

[0092] On the other hand, the present application further provides a computer-readable storage medium, and when the program is executed by a processor, the above-mentioned coal mine intelligent scheduling method based on voice recognition is implemented.

[0093] It can be understood that by performing noise reduction processing on the voice scheduling audio through wavelet transform, the recognition accuracy and reliability of the voice recognition system in a complex noise environment are improved, ensuring that voice scheduling can proceed smoothly in a high-noise environment such as a coal mine. Through semantic recognition of the signal to be recognized and comparison with the scheduling scheme database, the voice command is automatically converted into a specific scheduling scheme, and the key scheduling index in the scheduling scheme is extracted. Further introducing the dynamic calculation of the current task index and the demand index, by determining the comparison between the schedulable index and the scheduling index, it is automatically judged whether the scheduling scheme needs to be adjusted. The adjustment mechanism not only enhances the flexibility and adaptability of the scheduling scheme, but also can optimize the production task in real time. The comparison of the historical adjustment scheme further improves the intelligent level of the scheduling decision, enabling the scheduling system to make more accurate scheduling decisions based on experience and real-time data in a changing production environment. Through the intelligent and automated scheduling adjustment method, the accuracy and response speed of coal mine production scheduling are improved, manual intervention is reduced, and work efficiency and safety are improved.

[0094] Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0095] The present application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of the flows and / or blocks in the flowchart and / or block diagram can also be implemented. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0096] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including instruction means, and the instruction means implements the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0097] These computer program instructions 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 generate a computer-implemented process. Therefore, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0098] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: the specific implementation manners of the present invention can still be modified or equivalently replaced, and any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the protection scope of the claims of the present invention.

Claims

1. A coal mine intelligent scheduling method based on speech recognition, characterized in that, Including: Collect voice dispatching audio, and perform noise reduction processing on the voice dispatching audio based on wavelet transform to obtain a signal to be recognized; Perform semantic recognition on the signal to be recognized, compare the recognition features with a dispatching scheme database, determine a voice dispatching scheme, and extract the dispatching index in the voice dispatching scheme; Collect the current task index and demand index, determine the current task scope index according to the current task index, and determine the schedulable index according to the demand index and the current task scope index; Compare the schedulable index with the dispatching index, and judge whether to adjust the voice dispatching scheme according to the comparison result; When it is determined to adjust the voice dispatching scheme, establish the demand index, dispatching index, and schedulable index as a feature set, compare the feature set with historical adjustment schemes, and adjust the voice dispatching scheme according to the comparison result to obtain a final voice dispatching scheme.

2. The coal mine intelligent scheduling method based on speech recognition according to claim 1, wherein When performing noise reduction processing on the voice dispatching audio based on wavelet transform to obtain a signal to be recognized, it includes: Perform discrete wavelet transform on the voice dispatching audio using Daubechies 4 wavelet, perform 4-layer wavelet decomposition on the voice dispatching audio to obtain approximation coefficients and detail coefficients: Use the soft threshold method to perform threshold denoising on the detail coefficients, and set the threshold to 2 times the standard deviation of the signal noise; Perform inverse wavelet transform using the denoised coefficients to reconstruct the denoised signal to obtain the signal to be recognized.

3. The coal mine intelligent scheduling method based on speech recognition according to claim 1, characterized in that, When performing semantic recognition on the signal to be recognized, comparing the recognition features with a dispatching scheme database, determining a voice dispatching scheme, and extracting the dispatching index in the voice dispatching scheme, it includes: Convert the signal to be recognized into text based on semantic recognition; Perform word segmentation on the recognized text, extract keywords, and the keywords include resource type, starting position, target position, and dispatching index; Compare the recognized semantic features with the dispatching scheme database through keyword similarity matching, select the scheme with the highest similarity as the voice dispatching scheme, and determine the dispatching index according to the voice dispatching scheme.

4. The coal mine intelligent scheduling method based on speech recognition according to claim 1, wherein, When judging whether to adjust the voice dispatching scheme according to the comparison result, it includes: Collect the normal operation records corresponding to the demand index, and extract the corresponding demand index range from the normal operation records; Compare the dispatching index with the demand index range, and generate the current task scope index according to the numerical size relationship, where the current task scope index includes a left boundary value and a right boundary value; Obtain the index difference between the current task index and the left boundary of the current task scope index, and use the index difference as the schedulable index; When the schedulable index is less than or equal to the dispatching index, it is determined to adjust the voice dispatching scheme; When the schedulable index is greater than the dispatching index, it is determined not to adjust the voice dispatching scheme.

5. The coal mine intelligent scheduling method based on speech recognition according to claim 4, wherein When comparing the feature set with historical adjustment schemes and adjusting the voice dispatching scheme according to the comparison result, it includes: When there is data in the historical adjustment plan whose similarity to the feature set is greater than the similarity threshold, adjust the scheduling index in the voice scheduling plan according to the historical adjustment coefficient corresponding to the historical adjustment plan; when there is no data in the historical adjustment plan whose similarity to the feature set is greater than the similarity threshold, obtain a similarity set according to the historical adjustment plan, and determine an adjustment coefficient according to the similarity set to adjust the scheduling index in the voice scheduling plan.

6. The coal mine intelligent scheduling method based on speech recognition according to claim 5, characterized in that, When adjusting the scheduling index in the voice scheduling plan according to the historical adjustment coefficient corresponding to the historical adjustment plan, it includes: When the data in the historical adjustment plan whose similarity to the feature set is greater than the similarity threshold is unique, adjust the scheduling index in the voice scheduling plan with the historical adjustment coefficient corresponding to the historical adjustment plan; When the data in the historical adjustment plan whose similarity to the feature set is greater than the similarity threshold is not unique, adjust the scheduling index in the voice scheduling plan with the average value of the historical adjustment coefficients corresponding to the historical adjustment plan.

7. The coal mine intelligent scheduling method based on speech recognition according to claim 5, characterized in that When obtaining a similarity set according to the historical adjustment plan and determining an adjustment coefficient according to the similarity set to adjust the scheduling index in the voice scheduling plan, it includes: Establish a similarity set for the data in the historical adjustment plan whose similarity to the feature set is greater than a% and less than the similarity threshold; Select the historical adjustment coefficient corresponding to the data with the maximum similarity in the similarity set as the basic adjustment coefficient, and correct the basic adjustment coefficient according to the historical adjustment coefficients of the remaining data whose similarity is greater than the similarity threshold, and adjust the scheduling index in the voice scheduling plan with the corrected adjustment coefficient; Among them, represents the adjusted coefficient after correction, represents the number of data in the similar set whose remaining similarity is greater than the similarity threshold, represents the i-th historical adjustment coefficient corresponding to the data in the similar set whose remaining similarity is greater than the similarity threshold, represents the basic adjustment coefficient.

8. An intelligent coal mine scheduling device based on speech recognition, which is used to apply the intelligent coal mine scheduling method based on speech recognition according to any one of claims 1-7, and is characterized in that, It includes: An acquisition module, configured to acquire voice scheduling audio, and perform noise reduction processing on the voice scheduling audio based on wavelet transform to obtain a signal to be recognized; An identification module, configured to perform semantic identification on the signal to be recognized, compare the identification features with a scheduling plan database to determine a voice scheduling plan, and extract the scheduling index in the voice scheduling plan; A processing module, configured to acquire a current task index and a demand index, determine a current task range index according to the current task index, and determine an adjustable scheduling index according to the demand index and the current task range index; A judgment module, configured to compare the adjustable scheduling index with the scheduling index, and judge whether to adjust the voice scheduling plan according to the comparison result; The judgment module is further configured to, when it is determined to adjust the voice scheduling plan, establish the demand index, the scheduling index, and the adjustable scheduling index as a feature set, compare the feature set with the historical adjustment plan, and adjust the voice scheduling plan according to the comparison result to obtain a final voice scheduling plan.

9. A coal mine intelligent scheduling device based on speech recognition, characterized in that, It includes: One or more processors; A storage device, configured to store one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the coal mine intelligent scheduling method based on voice recognition as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the coal mine intelligent scheduling method based on speech recognition as described in any one of claims 1-7.

Images