Smart home intelligent control method, system and storage medium based on voice recognition

Abstract

The present invention discloses a smart home intelligent control method, system and storage medium based on voice recognition, and relates to the field of smart home technology. Firstly, a built-in timing sensor is used to collect response state information of smart devices in a voice recognition control process, and after feature extraction, a control response state information data set is constructed; according to the response state information in the relevant voice recognition control process, a voice recognition delay response coefficient Xyc is constructed, and if the voice recognition delay response coefficient Xyc exceeds a threshold, a delay warning instruction is issued externally; after receiving the delay warning instruction, data recognition and monitoring are performed on context noise conditions, network transmission speed and equipment load status, and a voice control influence coefficient Xyx is constructed. Combined with a trained voice control state prediction model, the obtained voice control evaluation index Zpg is compared and analyzed with a preset evaluation threshold P, and the corresponding level control instruction is obtained and executed.

Description

Technical Field

[0001] The present invention relates to the field of smart home technology, and in particular to a smart home intelligent control method, system and storage medium based on voice recognition. Background Art

[0002] With the advancement of technology and changes in lifestyle, smart home systems have become an important part of modern families. Smart homes provide a convenient and comfortable lifestyle by integrating various smart devices, such as smart lighting, smart temperature control and smart security. These devices are managed through a central control system, allowing users to control various devices in their homes through mobile phone applications and voice assistants. In particular, smart home control based on voice recognition achieves seamless interaction through voice commands, improving the convenience of operation and user experience. Through integration with smart voice assistants, users can control and manage devices with voice commands alone without manual operation.

[0003] In a Chinese invention patent with application publication number CN113314119B, a voice recognition smart home control method and device are disclosed. The control method receives a voice to be recognized from a second device through a first device, and the second device is a device that supports collecting user voices; recognizes the voice to be recognized to obtain a target text, which is a text corresponding to the voice of a target speaker in the voice to be recognized; determines a target device and a target operation instruction based on the target text; and sends a target operation instruction to the target device.

[0004] The above method can improve the accuracy of speech recognition by having the first device perform speech recognition on the speech to be recognized collected by the second device, and send the recognized target operation instructions to the target device through the first device, thereby enabling voice control of a smart home that does not have a speech recognition module to be realized remotely.

[0005] However, although the above-mentioned voice recognition smart home control method can realize voice control of smart homes that do not have voice recognition modules at the remote end, it still faces some challenges and shortcomings in practical applications; due to the long-distance signal transmission between multiple devices, in practical applications, due to the influence of environmental noise, network instability and device load conditions, the control response of voice recognition will have a certain delay, resulting in users having to wait for a long time to see the response results of the device after issuing voice commands, affecting the user's operating experience. Summary of the invention

[0006] In view of the deficiencies in the prior art, the present invention provides a smart home intelligent control method, system and storage medium based on voice recognition, which solves the problems in the above-mentioned background technology.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: a smart home intelligent control system based on voice recognition, including a response data acquisition module, a response analysis module, a voice control evaluation module and a dynamic control module;

[0008] The response data acquisition module is used to collect the response state information of the intelligent device during the speech recognition control process using the built-in timing sensor, and after feature extraction, to construct a control response state information data set;

[0009] The response analysis module is used to construct a speech recognition delay response coefficient Xyc according to the response state information in the relevant speech recognition control process, and if the speech recognition delay response coefficient Xyc exceeds a threshold, a delay warning instruction is issued externally;

[0010] The voice control evaluation module is used to perform data recognition and monitoring on the context noise, network transmission speed and device load status after receiving the delay warning instruction, obtain a voice control status data set after data preprocessing and feature extraction, construct a voice control influence coefficient Xyx according to the voice control status data set, and obtain a voice control evaluation index Zpg by fitting in combination with the trained voice control status prediction model;

[0011] The dynamic control module is used to pre-set an evaluation threshold P, and compare and analyze it with the voice control evaluation index Zpg, obtain corresponding level control instructions and execute them.

[0012] Preferably, the response data acquisition module is used to collect the response status information of the intelligent device during the voice recognition control process using a built-in timing sensor, wherein the response status information includes the time from receiving the voice signal to completing the signal recognition and conversion, the transmission time required for the voice signal when uploading and downloading, and the time it takes for the appliance to receive the instruction and complete the related execution operation; after feature extraction, a control response status information data set is constructed, and the control response status information data set includes the voice signal processing time Syc, the voice signal transmission time Scs, and the instruction execution action time Szx.

[0013] Preferably, the response analysis module includes a response analysis unit and a delay warning unit;

[0014] The response analysis unit is used to construct a speech recognition delay response coefficient Xyc according to the response state information in the relevant speech recognition control process. The speech recognition delay response coefficient Xyc is obtained by the following formula:

[0015]

[0016] In the formula, Syc represents the duration of voice signal processing, Scs represents the duration of voice signal transmission, and Szx represents the duration of instruction execution. It is the input of the logarithmic function, which represents the excessive impact of the voice signal processing time on the total time compared with the voice signal transmission time and the instruction execution time.

[0017] Preferably, the delay warning unit is used to compare and analyze the speech recognition delay response coefficient Xyc with a preset threshold value to preliminarily determine whether the response state in the current speech recognition control process is in a delayed response state. The specific comparison content is as follows:

[0018] If the speech recognition delayed response coefficient Xyc exceeds the threshold, it is preliminarily determined that the response state in the current speech recognition control process is in a delayed response state, and a delay warning instruction will be issued to the outside for further analysis;

[0019] If the voice recognition delayed response coefficient Xyc does not exceed the threshold, it is preliminarily determined that the response state in the current voice recognition control process is in a normal response state, and no additional warning instruction is issued at this time.

[0020] Preferably, the voice control evaluation module includes a voice control monitoring unit, a load analysis unit, a control state analysis unit and a voice control evaluation unit;

[0021] The voice control monitoring unit is used to perform data identification and monitoring on the context noise situation, network transmission speed and equipment load status after receiving the delay warning instruction, obtain voice control status data information, and pre-process the corresponding data information in the voice control status data information. The pre-processing includes cleaning the voice control status data information, removing invalid data, duplicate data and abnormal values, using statistical methods to detect and process abnormal values, and using smoothing methods to eliminate noise and random fluctuations to make the data more stable; and after feature extraction, a voice control status data set is constructed, and the voice control status data set includes environmental noise intensity Vqd, network transmission speed Vcs, CPU usage Vsy, memory occupancy Vnc and data read and write rate Vdx.

[0022] Preferably, the load analysis unit analyzes the load state of the device during voice control based on the voice control state data set, and obtains the device load coefficient Xfz after dimensionless processing; the device load coefficient Xfz is obtained by the following formula:

[0023]

[0024] In the formula, Vsy represents the CPU usage, Vnc represents the memory occupancy, and Vdx represents the data read and write rate. It is expressed as the synergy between the CPU usage Vsy and the memory occupancy Vnc. , and are all expressed as weight values.

[0025] Preferably, the control state analysis unit is used to associate the device load coefficient Xfz with the environmental noise intensity Vqd and the network transmission speed Vcs, and obtain the voice control influence coefficient Xyx after dimensionless processing. The voice control influence coefficient Xyx is obtained by the following formula:

[0026]

[0027] In the formula, , and They are respectively represented as weight values ​​of the equipment load factor Xfz, the environmental noise intensity Vqd and the network transmission speed Vcs;

[0028] The speech control evaluation unit inputs the speech recognition delay response coefficient Xyc and the speech control influence coefficient Xyx into the trained speech control state prediction model to obtain the speech control evaluation index Zpg by fitting. The speech control evaluation index Zpg is obtained by the following formula:

[0029]

[0030] In the formula, and They are respectively represented as the weight values ​​of the speech recognition delayed response coefficient Xyc and the speech control influence coefficient Xyx, and A is represented as a correction constant.

[0031] Preferably, the dynamic control module includes a determination unit and a control unit;

[0032] The determination unit is used to compare the voice control evaluation index Zpg with the preset evaluation threshold P to obtain the corresponding level control instruction, and the specific contents are as follows:

[0033] If the voice control evaluation index Zpg ≥ the evaluation threshold P, it indicates that the current voice recognition control process is determined to be in an abnormal delayed response state, and a first-level warning signal is generated;

[0034] If the voice control evaluation index Zpg is less than the evaluation threshold P, it indicates that the current voice recognition control process is determined to be in a normal delayed response state, and a secondary warning signal is generated;

[0035] The control unit is used to receive and generate a first-level warning signal and a second-level warning signal, and to make and execute corresponding level control instructions, the specific contents of which are as follows:

[0036] When a first-level warning signal is received, a first-level control instruction is generated and executed. The execution content is: triggering device load control instructions, including canceling background running tasks, optimizing network transmission speed by adjusting network bandwidth allocation and switching network connections, triggering optimization operations on the voice recognition model, including updating the voice recognition model version; starting the noise reduction function of the microphone; sending a notification to the user, informing that there is a delay problem in the current voice recognition control process, and the user needs to try the control operation again and reset the voice instruction later;

[0037] When a level-2 warning signal is received, a level-2 control instruction is generated and executed. The execution content is: continue to monitor the delay status of voice recognition, and record and store relevant data information to provide data support for subsequent analysis, and remind users through voice to continue normal use without additional operation.

[0038] A smart home intelligent control method based on voice recognition includes the following steps:

[0039] S1. Use the built-in timing sensor to collect the response state information of the intelligent device during the speech recognition control process, and after feature extraction, construct a control response state information dataset;

[0040] S2. Constructing a speech recognition delay response coefficient Xyc according to the response status information in the relevant speech recognition control process. If the speech recognition delay response coefficient Xyc exceeds a threshold, issuing a delay warning instruction.

[0041] S3. After receiving the delay warning instruction, the context noise, network transmission speed and equipment load status are identified and monitored, and after data preprocessing and feature extraction, a voice control status data set is obtained. The voice control influence coefficient Xyx is constructed according to the voice control status data set, and the voice control evaluation index Zpg is obtained by fitting in combination with the trained voice control status prediction model;

[0042] S4. Preset an evaluation threshold P, compare and analyze it with the voice control evaluation index Zpg, obtain and execute the corresponding level control instruction.

[0043] A smart home storage medium based on voice recognition, on which a computer program is stored, comprises the following steps: the computer program stores computer instructions, and when the instructions are executed by a processor, the smart home intelligent control system based on voice recognition is implemented.

[0044] The present invention provides a smart home intelligent control method, system and storage medium based on voice recognition, which have the following beneficial effects:

[0045] (1) By collecting and analyzing the response status information of voice signal processing, transmission and command execution, the voice recognition delay response coefficient Xyc is first constructed, and an early warning is issued when a delay is found to ensure that real-time adjustments can be made to deal with possible delay problems. Furthermore, by evaluating the environmental noise, network transmission speed and device load status, the voice control influence coefficient Xyx is obtained, and the voice control evaluation index Zpg is generated in combination with the trained prediction model. The result is then compared with the set threshold, and the corresponding control instructions are automatically generated and executed, thereby optimizing system performance and improving response speed and operation accuracy. This process not only makes the voice recognition control of smart home devices more stable and reliable, but also improves the user experience.

[0046] (2) The built-in timing sensor of the response data acquisition module accurately collects multiple key time parameters in the voice recognition control process to ensure comprehensive monitoring of the voice control response. Combined with the relevant formula of the delay response coefficient Xyc in the response analysis module, it can quantify each link of the voice recognition delay. By comparing the calculated delay response coefficient Xyc with the preset threshold, it is determined whether the voice control system is in a delayed state. When a delay problem is detected, the system can automatically issue a warning instruction to prompt further analysis and adjustment; in the absence of delay, the system maintains a normal state without generating additional interference. This process improves the response efficiency and stability of smart home voice control and ensures that potential delay problems can be quickly handled.

[0047] (3) Through the judgment unit and control unit in the dynamic control module, the delay status in the current voice recognition control process can be judged according to the change of the voice control evaluation index Zpg, accurately divided into normal and abnormal delay response states, and executed corresponding level control instructions. This design ensures the stability and flexibility of the smart home voice control system, reduces the device interruption caused by normal delay response, and ensures the efficient operation of the system in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is a block diagram of a smart home intelligent control system based on voice recognition in the present invention;

[0049] Figure 2 The present invention is a schematic diagram of the steps of a smart home intelligent control method based on voice recognition. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] Example 1

[0052] See also Figure 1 , the present invention provides a smart home intelligent control system based on voice recognition, including a response data acquisition module, a response analysis module, a voice control evaluation module and a dynamic control module;

[0053] The response data acquisition module is used to collect the response state information of the intelligent device during the speech recognition control process using the built-in timing sensor, and after feature extraction, to construct a control response state information data set;

[0054] The response analysis module is used to construct a speech recognition delay response coefficient Xyc according to the response state information in the relevant speech recognition control process, and if the speech recognition delay response coefficient Xyc exceeds a threshold, a delay warning instruction is issued externally;

[0055] The voice control evaluation module is used to perform data recognition and monitoring on the context noise, network transmission speed and device load status after receiving the delay warning instruction, obtain a voice control status data set after data preprocessing and feature extraction, construct a voice control influence coefficient Xyx according to the voice control status data set, and obtain a voice control evaluation index Zpg by fitting in combination with the trained voice control status prediction model;

[0056] The dynamic control module is used to pre-set an evaluation threshold P, and compare and analyze it with the voice control evaluation index Zpg, obtain corresponding level control instructions and execute them.

[0057] In this embodiment, the delay monitoring and response control in the process of smart home voice recognition control are realized through the mutual cooperation of the response data acquisition module, the response analysis module, the voice control evaluation module and the dynamic control module. First, the response status information of the smart device is collected in real time by using the built-in timing sensor, and a control response status information data set is constructed, which provides a reliable data basis for subsequent analysis. Secondly, the response analysis module can determine whether there is a delay problem by calculating the voice recognition delay response coefficient Xyc. Once the delay exceeds the preset threshold, an early warning instruction is immediately issued to ensure that the delay problem can be quickly identified. The voice control evaluation module monitors the multi-dimensional status of context noise, network transmission speed and device load in real time, and further constructs the voice control influence coefficient Xyx through data preprocessing and feature extraction. Combined with the trained voice control state prediction model, the system can dynamically fit the voice control evaluation index Zpg to provide an overall state evaluation of the response of the smart home intelligent control system based on voice recognition. Finally, through the comparison analysis of the evaluation threshold P of the dynamic control module, the corresponding level control instruction is generated and executed to ensure the stable operation and response optimization of the smart home system in a complex environment, improve the user experience, and reduce the adverse effects of delay problems on smart home control.

[0058] Example 2

[0059] Please refer to Figure 1 Specifically: the response data acquisition module is used to use the built-in timing sensor to collect the response status information of the intelligent device during the voice recognition control process, wherein the response status information includes the time from receiving the voice signal to completing the signal recognition and conversion, the transmission time required for the voice signal when uploading and downloading, and the time for the appliance to receive the instruction and complete the relevant execution operation; after feature extraction, a control response status information data set is constructed, and the control response status information data set includes the voice signal processing time Syc, the voice signal transmission time Scs and the instruction execution action time Szx.

[0060] Specifically, the response analysis module includes a response analysis unit and a delay warning unit;

[0061] The response analysis unit is used to construct a speech recognition delay response coefficient Xyc according to the response state information in the relevant speech recognition control process. The speech recognition delay response coefficient Xyc is obtained by the following formula:

[0062]

[0063] In the formula, Syc represents the duration of voice signal processing, Scs represents the duration of voice signal transmission, and Szx represents the duration of instruction execution. It is the input of the logarithmic function, which represents the excessive impact of the voice signal processing time on the total time compared with the voice signal transmission time and the instruction execution time.

[0064] Specifically, the delay warning unit is used to compare and analyze the speech recognition delay response coefficient Xyc with a preset threshold value to preliminarily determine whether the response state in the current speech recognition control process is in a delayed response state. The specific comparison content is as follows:

[0065] If the speech recognition delayed response coefficient Xyc exceeds the threshold, it is preliminarily determined that the response state in the current speech recognition control process is in a delayed response state, and a delay warning instruction will be issued to the outside for further analysis;

[0066] If the voice recognition delayed response coefficient Xyc does not exceed the threshold, it is preliminarily determined that the response state in the current voice recognition control process is in a normal response state, and no additional warning instruction is issued at this time.

[0067] In this embodiment, the built-in timing sensor of the response data acquisition module is used to accurately collect multiple key duration parameters in the voice recognition control process, including the voice signal processing duration Syc, the voice signal transmission duration Scs and the instruction execution action duration Szx, to ensure comprehensive monitoring of the voice control response. Combined with the relevant formula of the delay response coefficient Xyc in the response analysis module, each link of the voice recognition delay can be quantified, especially by introducing the logarithmic function to evaluate the impact of the voice signal processing duration Syc in the overall duration. The delay warning unit compares the calculated delay response coefficient Xyc with the preset threshold to determine whether the voice control system is in a delayed state. When a delay problem is detected, the system can automatically issue a warning instruction to prompt further analysis and adjustment. In the absence of delay, the system maintains a normal state without generating additional interference. This process improves the response efficiency and stability of smart home voice control, ensures that potential delay problems can be quickly handled, and thus improves user experience.

[0068] Example 3

[0069] Please refer to Figure 1 ,Specifically: the voice control evaluation module includes a voice control monitoring unit, a load analysis unit, a control state analysis unit and a voice control evaluation unit;

[0070] The voice control monitoring unit is used to, after receiving the delay warning instruction, perform data identification and monitoring on the context noise situation, network transmission speed and equipment load status through noise sensors, network speed test tools, system monitoring tools and disk performance test tools, obtain voice control status data information, and pre-process the corresponding data information in the voice control status data information. The pre-processing includes cleaning the voice control status data information, removing invalid data, duplicate data and abnormal values, using statistical methods to detect and process abnormal values, and using smoothing methods to eliminate noise and random fluctuations to make the data more stable; and after feature extraction, construct a voice control status data set, which includes environmental noise intensity Vqd, network transmission speed Vcs, CPU usage Vsy, memory occupancy Vnc and data read and write rate Vdx.

[0071] In this embodiment, after receiving the delay warning instruction, the voice control monitoring unit can perform real-time data identification and monitoring of the environmental noise conditions, network transmission speed and equipment load status, so as to grasp the factors affecting voice control. Through preprocessing steps, including data cleaning, removal of invalid and abnormal data, and use of statistical methods to process abnormal values, the system improves the accuracy and reliability of the data. At the same time, the application of the smoothing processing method further eliminates the noise and random fluctuations in the data, making the data more stable and ensuring the accuracy of subsequent analysis. After feature extraction, a voice control status data set including environmental noise intensity Vqd, network transmission speed Vcs, CPU usage Vsy, memory occupancy Vnc and data read and write rate Vdx parameters is constructed, which provides high-quality basic data support for subsequent control optimization and system evaluation. This optimization process ensures the adaptability and response efficiency of the voice control status in complex environments.

[0072] Example 4

[0073] Please refer to Figure 1 Specifically: the load analysis unit analyzes the load state of the device during the voice control process based on the voice control state data set, and obtains the device load coefficient Xfz after dimensionless processing; the device load coefficient Xfz is obtained by the following formula:

[0074]

[0075] In the formula, Vsy represents the CPU usage, Vnc represents the memory occupancy, and Vdx represents the data read and write rate. It is expressed as the synergy between the CPU usage Vsy and the memory occupancy Vnc. , and are all expressed as weight values.

[0076] It should be noted that the synergistic effect refers to the situation where the two influence each other and work together during the operation of the device, which means that when the CPU and memory work together, it is not just a simple superposition of their respective effects, but the cooperation of the two will produce additional effects or influences. Both are the core resources for smart devices to perform tasks. The CPU is responsible for processing and calculating instructions, while the memory is responsible for storing temporary data and providing space for the CPU to access data. If the CPU usage rate is high but the memory is insufficient, the overall performance of the device will decline, and vice versa. If the usage status between the two can be balanced and optimized, the operating efficiency of the device can be improved.

[0077] Specifically, the control state analysis unit is used to associate the device load coefficient Xfz with the environmental noise intensity Vqd and the network transmission speed Vcs, and obtain the voice control influence coefficient Xyx after dimensionless processing. The voice control influence coefficient Xyx is obtained by the following formula:

[0078]

[0079] In the formula, , and They are respectively represented as weight values ​​of the equipment load factor Xfz, the environmental noise intensity Vqd and the network transmission speed Vcs;

[0080] The speech control evaluation unit inputs the speech recognition delay response coefficient Xyc and the speech control influence coefficient Xyx into the trained speech control state prediction model to obtain the speech control evaluation index Zpg by fitting. The speech control evaluation index Zpg is obtained by the following formula:

[0081]

[0082] In the formula, and They are respectively represented as the weight values ​​of the speech recognition delayed response coefficient Xyc and the speech control influence coefficient Xyx, and A is represented as a correction constant.

[0083] In this embodiment, through the efficient collaboration of the load analysis unit, the control state analysis unit and the voice control evaluation unit, the evaluation and optimization of the device load and voice recognition delay in the voice control process are realized. First, the load analysis unit analyzes the load state of the device in real time when performing the voice control task based on the voice control state data set, and obtains the device load coefficient Xfz after dimensionless processing. The load state of the device is analyzed by comprehensively considering the CPU usage rate Vsy, the memory occupancy rate Vnc and the data read and write rate Vdx. Secondly, the control state analysis unit correlates the device load coefficient Xfz with the environmental noise intensity Vqd and the network transmission speed Vcs, and obtains the voice control influence coefficient Xyx through dimensionless processing. In this way, the control efficiency and responsiveness of the device in a complex environment are evaluated. Finally, the voice control evaluation unit inputs the voice recognition delay response coefficient Xyc and the voice control influence coefficient Xyx into the trained prediction model to fit and generate the voice control evaluation index Zpg. The voice control evaluation index Zpg comprehensively reflects the influence of multiple factors such as system delay, environmental interference and device load during the voice control process. Through this evaluation mechanism, the real-time operating status of the smart home voice control can be judged, and timely adjustments can be made when abnormalities are found, thereby improving the system's stability, response speed and user experience, ensuring that the smart home system can operate stably and efficiently in complex and changeable actual usage scenarios, and meeting users' needs for instant voice control.

[0084] Example 5

[0085] Please refer to Figure 1 , specifically: the dynamic control module includes a determination unit and a control unit;

[0086] The determination unit is used to compare the voice control evaluation index Zpg with the preset evaluation threshold P to obtain the corresponding level control instruction, and the specific contents are as follows:

[0087] If the voice control evaluation index Zpg ≥ the evaluation threshold P, it indicates that the current voice recognition control process is determined to be in an abnormal delayed response state, and a first-level warning signal is generated;

[0088] If the voice control evaluation index Zpg is less than the evaluation threshold P, it indicates that the current voice recognition control process is determined to be in a normal delayed response state, and a secondary warning signal is generated;

[0089] The control unit is used to receive and generate a first-level warning signal and a second-level warning signal, and to make and execute corresponding level control instructions, the specific contents of which are as follows:

[0090] When a first-level warning signal is received, a first-level control instruction is generated and executed. The execution content is: triggering device load control instructions, including canceling background running tasks, optimizing network transmission speed by adjusting network bandwidth allocation and switching network connections, triggering optimization operations on the voice recognition model, including updating the voice recognition model version; starting the noise reduction function of the microphone; sending a notification to the user, informing that there is a delay problem in the current voice recognition control process, and the user needs to try the control operation again and reset the voice instruction later;

[0091] When a level-2 warning signal is received, a level-2 control instruction is generated and executed. The execution content is: continue to monitor the delay status of voice recognition, and record and store relevant data information to provide data support for subsequent analysis, and remind users through voice to continue normal use without additional operation.

[0092] In this embodiment, through the determination unit and the control unit in the dynamic control module, the delay status in the current voice recognition control process can be determined according to the change of the voice control evaluation index Zpg, and accurately divided into normal and abnormal delay response states. Specifically, when the voice control evaluation index Zpg is greater than or equal to the preset evaluation threshold P, a first-level warning signal is generated to trigger a series of optimization measures, such as adjusting the device load, canceling background tasks, optimizing network bandwidth allocation and connection, updating the voice recognition model version, and starting the noise reduction function of the microphone. This process not only improves the operating efficiency of the device, but also notifies the user of the current voice recognition control delay problem by notifying the user and allowing the user to make adjustments. Use behavior to avoid further delays and control failures. At the same time, when the voice control evaluation index Zpg is less than the preset evaluation threshold P, the system generates a secondary warning signal, continues to monitor the delay status of voice recognition, and records relevant data information for subsequent analysis. During this process, the user does not need to intervene in any way. The system confirms through voice prompts that the current delay status is within an acceptable range, ensuring the continuity of the system and the smoothness of the user experience. This design can ensure the stability and flexibility of the smart home voice control system and reduce the interruption of equipment caused by normal delayed response. The intelligent response capability of this module improves the self-regulation performance of the system and ensures the efficient operation of the system in complex environments.

[0093] Example 6

[0094] Please refer to Figure 1 and Figure 2 Specifically: A smart home intelligent control method based on voice recognition, comprising the following steps:

[0095] S1. Use the built-in timing sensor to collect the response state information of the intelligent device during the speech recognition control process, and after feature extraction, construct a control response state information dataset;

[0096] S2. Constructing a speech recognition delay response coefficient Xyc according to the response status information in the relevant speech recognition control process. If the speech recognition delay response coefficient Xyc exceeds a threshold, issuing a delay warning instruction.

[0097] S3. After receiving the delay warning instruction, the context noise, network transmission speed and equipment load status are identified and monitored, and after data preprocessing and feature extraction, a voice control status data set is obtained. The voice control influence coefficient Xyx is constructed according to the voice control status data set, and the voice control evaluation index Zpg is obtained by fitting in combination with the trained voice control status prediction model;

[0098] S4. Preset an evaluation threshold P, compare and analyze it with the voice control evaluation index Zpg, obtain and execute the corresponding level control instruction.

[0099] A smart home storage medium based on voice recognition, on which a computer program is stored, comprises the following steps: the computer program stores computer instructions, and when the instructions are executed by a processor, the smart home intelligent control system based on voice recognition is implemented.

[0100] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A smart home intelligent control system based on voice recognition, characterized in that: It includes a response data acquisition module, a response analysis module, a voice control evaluation module and a dynamic control module; The response data acquisition module is used to collect the response state information of the intelligent device during the speech recognition control process using the built-in timing sensor, and after feature extraction, to construct a control response state information data set; The response data acquisition module is used to collect the response status information of the intelligent device during the voice recognition control process using the built-in timing sensor, wherein the response status information includes the time from receiving the voice signal to completing the signal recognition and conversion, the transmission time required for the voice signal when uploading and downloading, and the time for the appliance to receive the instruction and complete the relevant execution operation; after feature extraction, a control response status information data set is constructed, and the control response status information data set includes the voice signal processing time Syc, the voice signal transmission time Scs, and the instruction execution action time Szx; The response analysis module is used to construct a speech recognition delay response coefficient Xyc according to the response state information in the relevant speech recognition control process, and if the speech recognition delay response coefficient Xyc exceeds a threshold, a delay warning instruction is issued externally; The response analysis module includes a response analysis unit and a delay warning unit; The response analysis unit is used to construct a speech recognition delay response coefficient Xyc according to the response state information in the relevant speech recognition control process. The speech recognition delay response coefficient Xyc is obtained by the following formula: In the formula, Syc represents the duration of voice signal processing, Scs represents the duration of voice signal transmission, and Szx represents the duration of instruction execution. is the input of the logarithmic function, which represents the excessive impact of the voice signal processing time compared to the voice signal transmission time and the command execution time on the total time; The voice control evaluation module is used to perform data recognition and monitoring on the context noise, network transmission speed and device load status after receiving the delay warning instruction, obtain a voice control status data set after data preprocessing and feature extraction, construct a voice control influence coefficient Xyx according to the voice control status data set, and obtain a voice control evaluation index Zpg by fitting in combination with the trained voice control status prediction model; The voice control evaluation module includes a voice control monitoring unit, a load analysis unit, a control state analysis unit and a voice control evaluation unit; The voice control state data set includes the environmental noise intensity Vqd, the network transmission speed Vcs, the CPU usage rate Vsy, the memory occupancy rate Vnc and the data read and write rate Vdx; The load analysis unit analyzes the load state of the device during the voice control process based on the voice control state data set, and obtains the device load coefficient Xfz after dimensionless processing; the device load coefficient Xfz is obtained by the following formula: In the formula, Vsy represents the CPU usage, Vnc represents the memory occupancy, and Vdx represents the data read and write rate. It is expressed as the synergy between the CPU usage Vsy and the memory occupancy Vnc. , and All are expressed as weight values; The control state analysis unit is used to associate the device load coefficient Xfz with the environmental noise intensity Vqd and the network transmission speed Vcs, and obtain the voice control influence coefficient Xyx after dimensionless processing. The voice control influence coefficient Xyx is obtained by the following formula: In the formula, , and They are respectively represented as weight values ​​of the equipment load factor Xfz, the environmental noise intensity Vqd and the network transmission speed Vcs; The speech control evaluation unit inputs the speech recognition delay response coefficient Xyc and the speech control influence coefficient Xyx into the trained speech control state prediction model to obtain the speech control evaluation index Zpg by fitting. The speech control evaluation index Zpg is obtained by the following formula: In the formula, and They are respectively represented as weight values ​​of the speech recognition delay response coefficient Xyc and the speech control influence coefficient Xyx, and A is represented as a correction constant; The dynamic control module is used to pre-set an evaluation threshold P, and compare and analyze it with the voice control evaluation index Zpg, obtain corresponding level control instructions and execute them.

2. According to claim 1, a smart home intelligent control system based on voice recognition is characterized in that: The delay warning unit is used to compare and analyze the speech recognition delay response coefficient Xyc with a preset threshold value to preliminarily determine whether the response state in the current speech recognition control process is in a delayed response state. The specific comparison content is as follows: If the speech recognition delayed response coefficient Xyc exceeds the threshold, it is preliminarily determined that the response state in the current speech recognition control process is in a delayed response state, and a delay warning instruction will be issued to the outside for further analysis; If the voice recognition delayed response coefficient Xyc does not exceed the threshold, it is preliminarily determined that the response state in the current voice recognition control process is in a normal response state, and no additional warning instruction is issued at this time.

3. According to claim 1, a smart home intelligent control system based on voice recognition is characterized in that: The voice control monitoring unit is used to perform data recognition and monitoring on the context noise situation, network transmission speed and equipment load status after receiving the delay warning instruction, obtain voice control status data information, and pre-process the corresponding data information in the voice control status data information. The pre-processing includes cleaning the voice control status data information, removing invalid data, duplicate data and outliers, using statistical methods to detect and process outliers, using smoothing methods to eliminate noise and random fluctuations, and making the data more stable; and after feature extraction, constructing a voice control status data set.

4. The smart home intelligent control system based on voice recognition according to claim 1, characterized in that: The dynamic control module includes a determination unit and a control unit; The determination unit is used to compare the voice control evaluation index Zpg with the preset evaluation threshold P to obtain the corresponding level control instruction, and the specific contents are as follows: If the voice control evaluation index Zpg ≥ the evaluation threshold P, it indicates that the current voice recognition control process is determined to be in an abnormal delayed response state, and a first-level warning signal is generated; If the voice control evaluation index Zpg is less than the evaluation threshold P, it indicates that the current voice recognition control process is determined to be in a normal delayed response state, and a secondary warning signal is generated; The control unit is used to receive and generate a first-level warning signal and a second-level warning signal, and to make and execute corresponding level control instructions, the specific contents of which are as follows: When a first-level warning signal is received, a first-level control instruction is generated and executed, and the execution content is: triggering device load control instructions, including canceling background running tasks, optimizing network transmission speed by adjusting network bandwidth allocation and switching network connections, and triggering optimization operations on the speech recognition model, including updating the speech recognition model version; Activate the noise reduction function of the microphone; send a notification to the user, informing him that there is a delay in the current voice recognition control process, and the user needs to try the control operation again and reset the voice command later; When a level-2 warning signal is received, a level-2 control instruction is generated and executed. The execution content is: continue to monitor the delay status of voice recognition, and record and store relevant data information to provide data support for subsequent analysis, and remind users through voice to continue normal use without additional operation.

5. A smart home intelligent control method based on voice recognition, used to implement a smart home intelligent control system based on voice recognition as described in any one of claims 1 to 4, characterized in that: The following steps are included: S1. Use the built-in timing sensor to collect the response state information of the intelligent device during the speech recognition control process, and after feature extraction, construct a control response state information dataset; S2. Constructing a speech recognition delay response coefficient Xyc according to the response status information in the relevant speech recognition control process. If the speech recognition delay response coefficient Xyc exceeds a threshold, issuing a delay warning instruction. S3. After receiving the delay warning instruction, the context noise, network transmission speed and equipment load status are identified and monitored, and after data preprocessing and feature extraction, a voice control status data set is obtained. The voice control influence coefficient Xyx is constructed according to the voice control status data set, and the voice control evaluation index Zpg is obtained by fitting in combination with the trained voice control status prediction model; S4. Preset an evaluation threshold P, compare and analyze it with the voice control evaluation index Zpg, obtain and execute the corresponding level control instruction.

6. A smart home storage medium based on speech recognition, on which a computer program is stored, characterized in that: The computer program stores computer instructions, and when the instructions are executed by a processor, a smart home intelligent control system based on voice recognition as described in any one of claims 1-4 is implemented.

Images