Sound automatic tuning method and system based on intelligent sensor

Abstract

The application discloses a sound automatic tuning method and system based on intelligent sensors, and relates to the technical field of sound control, comprising the following steps: collecting sound field data of each intelligent sensor in a target sound field and sound system tuning parameters, dividing each intelligent sensor into zones according to the response of each channel, and obtaining sound field results corresponding to each zone; comparing the sound field results with contrast results after tuning is completed, calculating the change and determining the change source; classifying and storing the change record, judging continuity and consistency, and generating an adjustment instruction; modifying the gain value, equalization parameter or delay parameter in the corresponding zone according to the adjustment instruction, and verifying the modified sound field results; the method can relatively reduce the dependence on artificial tuning judgment, and improve the continuity, stability and consistency of tuning decision under dynamic sound field conditions.

Description

Technical Field

[0001] This invention relates to the field of audio control technology, specifically to an automatic audio tuning method and system based on intelligent sensors. Background Technology

[0002] In the application scenarios of digital professional audio equipment, the tuning process usually requires comprehensive adjustments to the frequency response, phase relationship, sound pressure level distribution, coverage uniformity, and feedback margin of the speaker system. Common technical approaches include single-point or multi-point measurement tuning based on test signals, equalization and delay correction based on digital signal processors, and semi-automatic tuning schemes combined with preset scene parameters. Although the above techniques can form a relatively usable set of system parameters in the early stages of setup, their applicability is generally based on the assumptions that the sound field conditions are relatively stable, the reference measurement points are representative, and the tuning target remains basically unchanged over a long period of time. However, in actual use, changes in sound absorption after the audience enters, the flow of people, and Temporary obstructions causing changes in the propagation path, changes in coupling between the stage monitor and the main amplification link, slight offsets in equipment installation angles or placements, crosstalk between adjacent amplification equipment, and changes in the target caused by program format switching can all cause the initial sound mixing results to continuously deviate from the actual needs. In this case, existing automatic sound mixing technology often struggles to accurately determine whether the sound field changes are continuous, whether they constitute a valid offset that needs correction, and within what range adjustments should be made. Therefore, the system usually still relies on experienced sound mixers to repeatedly compare, judge, and correct based on on-site listening, measurement results, and the intended use, in order to avoid local compensation disrupting the overall balance, short-term disturbances triggering incorrect adjustments, or multi-region parameter linkage instability.

[0003] It is evident that the current problem lies not only in insufficient tuning accuracy, but also in the fact that traditional audio tuning relies heavily on manual operation and judgment, making it difficult to make stable, continuous, and consistent tuning decisions under dynamic sound field conditions. Summary of the Invention

[0004] The purpose of this invention is to provide an automatic audio tuning method and system based on intelligent sensors, which can automatically identify and adjust the sound field changes in different zones, reduce the degree of human intervention, and improve the automation level, response time and adjustment stability of audio tuning.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] Automatic audio tuning methods based on smart sensors include:

[0007] S1. Acquire the sound field data collected by each intelligent sensor deployed in the target sound reinforcement site and the tuning parameters of the audio system. Divide the sound field data into partitions according to the response magnitude of each intelligent sensor to each channel, and output the sound field results corresponding to each partition.

[0008] S2. Compare the sound field results corresponding to each zone with the control results. The control results are the sound field results of the corresponding zone after the tuning is completed. Calculate the change amount of each zone, and determine the source of change based on the change amount during the continuous sampling period. Output the change record corresponding to each zone.

[0009] S3. Classify and store the change records corresponding to each partition according to the source of change, and record the partition, time period, direction of change and adjustment range corresponding to each change record, and output the updated record set;

[0010] S4. Perform continuity and consistency checks on the change records belonging to the same change source in the updated record set, determine the target record that meets the adjustment conditions, and output the adjustment instruction.

[0011] S5. Based on the adjustment command, determine the target parameters corresponding to the target record from the tuning parameters, modify only the gain value, equalization parameter or delay parameter corresponding to the adjustment range in the corresponding partition of the target record, and output the updated tuning parameters.

[0012] S6. Based on the updated tuning parameters and the sound field data subsequently collected by the smart sensor, verify the sound field results after parameter modification, and when the corresponding sound field change disappears or the subsequent sound field data indicates that the original source of change is not valid, cancel or pause the local adjustment corresponding to the target record, and output the automatic tuning result.

[0013] Preferably, S1 includes:

[0014] S1.1 Read the installation location, acquisition time, sound pressure level of each frequency band and arrival time of direct sound collected by each smart sensor as sound field data; read the channel identifier, gain value, equalization parameter and delay parameter of each channel of the audio system as tuning parameters; output the sound field data and tuning parameters aligned with the acquisition time.

[0015] S1.2 Calculate the response size for each channel of each smart sensor. The response size is determined according to the following rules: Subtract the delay parameter of the corresponding channel from the arrival time of the direct sound of the smart sensor to obtain the arrival time difference. Compare the direction of change of the sound pressure level of each frequency band of the smart sensor with the direction of change of the gain value and equalization parameter of the corresponding channel band by band to obtain the number of frequency bands with the same direction. First sort them by the absolute value of the arrival time difference from smallest to largest. Then, when the absolute values ​​of the arrival time differences are the same, sort them by the number of frequency bands with the same direction from most to least. Identify the channel with the first ranking as the channel to which the smart sensor belongs, and divide the sound field data with the same channel into the same partition.

[0016] S1.3. Summarize the sound field data in each zone according to the acquisition time, take the average value of the sound pressure level of each frequency band and the average value of the arrival time of the direct sound to obtain the sound field results corresponding to each zone; when the corresponding home channel of the same smart sensor is inconsistent in three adjacent acquisition times, retain the channel identifier that appears most frequently as the home channel of the smart sensor, and re-divide the zones according to the retained home channel and update the sound field results corresponding to each zone.

[0017] Preferably, S2 includes:

[0018] S2.1 Read the sound field results corresponding to each partition and the corresponding comparison results of the partition. Subtract the sound pressure level of each frequency band in the comparison results from the sound pressure level of each frequency band in the sound field results, and subtract the arrival time of the direct sound in the comparison results from the arrival time of the direct sound in the sound field results to obtain the sound pressure level difference and arrival time difference of each partition in each frequency band, which are used as the change amount corresponding to the partition.

[0019] S2.2. Arrange the changes in the same zone during continuous sampling periods in the order of sampling time, and compare the signs of the sound pressure level difference and the arrival time difference between two adjacent sampling periods for each period. When the signs of the sound pressure level difference are continuously the same and the arrival time difference remains unchanged, the source of change is determined to be sound absorption change. When the signs of the sound pressure level difference are continuously the same and the arrival time difference changes continuously in the same direction, the source of change is determined to be propagation change. When the signs of the sound pressure level difference change before and after or the arrival time difference changes before and after, the source of change is determined to be transient interference, and the corresponding source of change for each zone is output.

[0020] S2.3 Write the partition identifier, acquisition time period, sound pressure level difference, arrival time difference, and change source into the corresponding change record; when the change source of the same partition is the same in two adjacent sampling time periods, write a continuous mark in the next change record; when the change source of the same partition is different in two adjacent sampling time periods, write a switching mark in the next change record, and output the change record corresponding to each partition.

[0021] Preferably, step S3 includes:

[0022] S3.1 Read the partition identifier, collection period, sound pressure level difference, arrival time difference and change source from the change records corresponding to each partition. Classify and write each change record according to the change source, partition identifier and collection period. Arrange the change records with the same partition identifier under the same change source in the order of collection period to obtain the record set.

[0023] S3.2 Determine the positive and negative directions of the sound pressure level difference for each change record in the record set for each frequency band. Write the positive and negative directions that occur most frequently as the direction of change, write the frequency bands where the sound pressure level difference is not zero as the adjustment range, and write the corresponding partition identifier, acquisition time period, direction of change and adjustment range into each change record. Output the updated record set.

[0024] Preferably, S4 includes:

[0025] S4.1 Read all change records with the same change source in the record set and arrange them in order of partition identifier and collection time period; compare the change direction and adjustment range of adjacent change records in the same partition. When the change direction and adjustment range are the same, connect the corresponding change records into continuous records in sequence. Otherwise, end the current continuous record and start connecting again.

[0026] S4.2 Count the number of collection periods included in each continuous record, and determine the continuous record with no less than two collection periods as the target record. Write the partition identifier, change direction and adjustment range in the target record into the adjustment instruction and output the adjustment instruction.

[0027] S4.3. For continuous records formed under the same source of change, continue to read the change records corresponding to the next collection period, and compare the change direction and adjustment range of the next change record with the last change record of the continuous record; when the change direction or adjustment range is different, write the continuous record into the target record set, and write the next change record into the first record of the new continuous record; otherwise, continue to write the next change record into the current continuous record.

[0028] S4.4 Generate adjustment instructions for each target record in the target record set, and write the partition identifier, change source, change direction, adjustment range and corresponding collection time period into the adjustment instructions; when there are two target records with different change sources in the same partition in the same collection time period, retain the adjustment instructions corresponding to the target record with more collection time periods and delete the other adjustment instructions;

[0029] S4.5 Read the target record corresponding to the output adjustment command, and continue to read the change record of the corresponding partition in the subsequent acquisition period; when the change direction and adjustment range of the subsequent change record are the same as the target record, append the subsequent change record to the target record and update the acquisition period in the corresponding adjustment command;

[0030] S4.6 When the direction of change of subsequent change records is opposite to that of the target record, or the adjustment range of subsequent change records is different from that of the target record, the append writing stops, the original adjustment instruction is written as a stop execution instruction, and the subsequent change records are written back to the record set as the new judgment starting point.

[0031] Preferably, S5 includes:

[0032] S5.1 Read the zone identifier, change source, change direction, and adjustment range from the adjustment command, and read the gain value, equalization parameter, and delay parameter of the corresponding channel from the tuning parameters according to the zone identifier; when the change source is sound absorption change, determine the equalization parameter and gain value corresponding to the adjustment range as the target parameter; when the change source is propagation change, determine the delay parameter and gain value corresponding to the adjustment range as the target parameter.

[0033] S5.2 Modify the target parameters one by one according to the direction of change; when the direction of change is increasing, decrease the gain value or equalization parameter of the corresponding frequency band within the adjustment range step by step, or decrease the delay parameter in the opposite direction of the arrival time difference; when the direction of change is decreasing, increase the gain value or equalization parameter of the corresponding frequency band within the adjustment range step by step, or increase the delay parameter in the same direction as the arrival time difference.

[0034] S5.3 Write the modified target parameters back to the tuning parameters of the corresponding partition, while keeping the tuning parameters of other partitions unchanged, and output the updated tuning parameters.

[0035] Preferably, S6 includes:

[0036] S6.1 Read the updated tuning parameters and the sound field data collected by the smart sensor in the subsequent acquisition period. Divide the sound field data into partitions according to the response of each smart sensor to each channel. Output the sound field results corresponding to each partition. Compare the sound field results corresponding to each partition with the corresponding partition's comparison results to obtain the amount of change and the source of change of each partition in the subsequent acquisition period.

[0037] S6.2. Compare the amount of change and the source of change of each partition in the subsequent collection period with the amount of change and the source of change corresponding to the target record; when the amount of change is zero, cancel the parameter modification corresponding to the target record; when the source of change is different from the source of change in the target record, pause the parameter modification corresponding to the target record and retain the parameter value before modification; when the amount of change decreases and the source of change remains unchanged, maintain the parameter modification corresponding to the target record.

[0038] S6.3 Write the parameter status after cancellation, pause or hold into the tuning parameters of the corresponding partition, and write the corresponding partition identifier, subsequent acquisition period, change amount, change source and parameter status into the automatic tuning result, and output the automatic tuning result.

[0039] An automatic audio tuning system based on intelligent sensors, used to implement the method, includes:

[0040] The data acquisition module acquires sound field data collected by various intelligent sensors deployed in the target sound reinforcement venue, as well as the tuning parameters of the audio system.

[0041] The partitioning module partitions the sound field data according to the response magnitude of each smart sensor to each channel and outputs the sound field results corresponding to each partition.

[0042] The change analysis module compares the sound field results corresponding to each zone with the control results, which are the sound field results of the corresponding zone after tuning. It calculates the change amount of each zone, determines the source of change based on the change amount during the continuous sampling period, and outputs the change record corresponding to each zone.

[0043] The record processing module categorizes and stores the change records corresponding to each partition according to the source of change, and records the partition, time period, direction of change and adjustment range corresponding to each change record, and outputs the updated record set;

[0044] The instruction generation module performs continuity and consistency checks on change records belonging to the same source of change in the updated record set, determines the target record that meets the adjustment conditions, and outputs the adjustment instruction.

[0045] The parameter adjustment module determines the target parameters corresponding to the target record from the tuning parameters according to the adjustment command, and only modifies the gain value, equalization parameter or delay parameter corresponding to the adjustment range in the corresponding partition of the target record, and outputs the updated tuning parameters.

[0046] The result verification module verifies the sound field results after parameter modification based on the updated tuning parameters and the sound field data subsequently collected by the smart sensor. When the corresponding sound field change disappears or the subsequent sound field data indicates that the original source of the change is not valid, the module cancels or pauses the local adjustment corresponding to the target record and outputs the automatic tuning result.

[0047] Compared with the prior art, the beneficial effects of the present invention are:

[0048] 1. By combining intelligent sensor acquisition, zone sound field comparison, continuous recording and judgment, and parameter feedback closed loop, the identification of changes in the dynamic sound field, adjustment decision and result verification are linked together. This can relatively reduce the dependence of traditional tuning on manual operation and judgment, and make the tuning decision under dynamic sound field conditions more stable, continuous and consistent.

[0049] 2. By using the intelligent sensors to calculate the response magnitude of each channel, the sound field data is divided into zones according to the real-time response relationship, and the corresponding sound field results are generated. Compared with the method of dividing the zone according to a fixed position, it can be closer to the actual sound reinforcement coverage relationship and improve the pertinence of subsequent change analysis and parameter adjustment.

[0050] 3. By comparing the current zone sound field results with the control results after tuning, and combining the continuity of changes in the amount of change within the continuous sampling period to distinguish sound absorption changes, propagation changes and instantaneous interference, the interference of single fluctuations on the judgment results can be relatively suppressed, which is conducive to improving the reliability of the source of change identification.

[0051] 4. Organize the change records according to the source of change, partition identifier and collection period, and add the change direction and adjustment range. This allows subsequent continuous judgment and adjustment instruction generation to be directly based on the structured records, which can reduce the repeated parsing of the original difference and relatively improve the coherence and execution efficiency of the adjustment process.

[0052] 5. Modifying only the gain, equalization, or delay parameters within the corresponding zone of the target record and the adjustment range, while keeping the tuning parameters of other zones unchanged, can reduce the cascading effect of single-zone adjustment on unrelated zones to a certain extent and improve the independence of zone adjustment;

[0053] 6. After parameter modification, continue to collect subsequent sound field data, and decide whether to keep, pause or cancel the corresponding parameter modification based on the change amount and the change source. It can roll back or stop the adjustment that is no longer suitable for the current sound field state, and alleviate the adjustment deviation problem caused by long-term parameter retention. Attached Figure Description

[0054] Figure 1 This is a flowchart of the method of the present invention;

[0055] Figure 2 This is a system block diagram of the present invention. Detailed Implementation

[0056] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0057] This invention provides an automatic sound tuning method and system based on intelligent sensors, applicable to automatic adjustment scenarios of sound systems in conference sound reinforcement, performance sound reinforcement, and venue sound reinforcement. The solution deploys multiple intelligent sensors within the target sound reinforcement area to collect sound field data from various locations. Combined with the sound system's tuning parameters, the sound field results for different zones are acquired and compared to further identify the amount and source of change in each zone. Based on this, the change records are categorized, stored, and their continuity is assessed. Corresponding adjustment commands are generated to modify the gain value, equalization parameters, or delay parameters corresponding to the adjustment range within the target zone. The sound field results after parameter modification are then verified, thereby achieving automatic tuning of the sound system in dynamic sound field environments and reducing reliance on manual operation and judgment.

[0058] Example

[0059] S1. Acquire the sound field data collected by each intelligent sensor deployed in the target sound reinforcement site and the tuning parameters of the audio system. Divide the sound field data into partitions according to the response magnitude of each intelligent sensor to each channel, and output the sound field results corresponding to each partition.

[0060] In this implementation process, step S1 involves correlating the sound field data collected by each smart sensor within the target sound reinforcement site with the tuning parameters effective at the same time for each channel of the audio system. Based on this, it determines the channel most likely to correspond to each smart sensor, and then merges the sound field data belonging to the same channel into the same zone, ultimately forming the sound field results corresponding to each zone. This process is not simply dividing zones according to fixed installation locations, but first utilizes the correspondence between the arrival time of direct sound and the channel delay parameters, and then combines the consistency between the direction of sound pressure level change in each frequency band and the direction of channel parameter change to complete the channel assignment determination; then, the data from multiple smart sensors are summarized by zone to obtain a sound field result that reflects the overall sound reinforcement status of the zone. After this processing, subsequent steps can directly compare and adjust based on the zone-level results, without having to process discrete sampling points one by one. This implementation process includes the following steps:

[0061] S1.1 Read the installation location, acquisition time, sound pressure level of each frequency band, and direct sound arrival time from each smart sensor as sound field data. Read the channel identifier, gain value, equalization parameter, and delay parameter of each channel of the audio system as tuning parameters. Output the sound field data and tuning parameters aligned with the acquisition time. In specific implementation, each smart sensor directly outputs its corresponding installation location, acquisition time, sound pressure level of each frequency band, and direct sound arrival time; the audio system's control device outputs the channel identifier, gain value, equalization parameter, and delay parameter effective at the corresponding time for each channel. Then, based on the acquisition time as the alignment basis, pair the sound field data corresponding to each acquisition time with the tuning parameters already effective at that acquisition time; when the update time of the tuning parameters is not completely consistent with the acquisition time, take the most recent effective tuning parameter before that acquisition time as the corresponding parameter. After pairing, sound field data and tuning parameters corresponding one-to-one with the acquisition time are formed, which serve as input for the next step of calculating the response size.

[0062] S1.2 Calculate the response magnitude for each channel of each smart sensor. The response magnitude is determined according to the following rules: Subtract the delay parameter of the corresponding channel from the arrival time of the direct sound of the smart sensor to obtain the arrival time difference. Compare the direction of change of the sound pressure level of each frequency band of the smart sensor with the direction of change of the corresponding channel's gain value and equalization parameter band by band to obtain the number of frequency bands with the same direction. First, sort them by the absolute value of the arrival time difference from smallest to largest. Then, when the absolute values ​​of the arrival time differences are the same, sort them by the number of frequency bands with the same direction from largest to smallest. Identify the channel with the highest ranking as the channel to which the smart sensor belongs, and classify the sound field data with the same channel as... Within the same partition; in specific implementation, all channels are sequentially traversed for each smart sensor, and the response magnitude between the smart sensor and each channel is calculated respectively; the arrival time difference is directly obtained by subtracting the delay parameter of the channel from the arrival time of the direct sound of the smart sensor; the smaller the absolute value of the arrival time difference, the closer the time characteristic of the direct sound received by the smart sensor is to the output delay of the channel; the direction of change of sound pressure level in each frequency band is obtained by comparing the sound pressure level of the corresponding frequency band of the same smart sensor at the current acquisition time with that at the previous acquisition time, if the current value is greater than that at the previous time, it is recorded as an increase, if it is less than that at the previous time, it is recorded as a decrease, and if they are the same, it is recorded as no change; the frequency band of the corresponding channel The direction of change is directly determined by the gain value and equalization parameters of the channel. An increase in gain indicates an overall increase, a decrease in gain indicates an overall decrease, and an increase in equalization parameters within the corresponding frequency band indicates an increase in that band, an decrease indicates a decrease, and no adjustment indicates no change. Subsequently, the direction of change of each frequency band of the smart sensor is compared band by band with the direction of change of the corresponding frequency band of the channel, and the number of frequency bands with the same direction is counted. After calculating all channels, they are first sorted by the absolute value of their arrival time difference. If multiple channels have the same absolute value of arrival time difference, the number of frequency bands with the same direction is compared among these channels, and the channel with more frequency bands with the same direction is ranked first. If they are still the same, the channel with the earlier channel identifier is retained. Finally, the channel identifier with the first order is determined as the channel to which the smart sensor belongs. For multiple smart sensors with the same channel, their corresponding sound field data are assigned to the same partition. The partitions obtained in this way are not pre-defined, but are formed based on the real-time response relationship, which better reflects the actual sound reinforcement coverage relationship. For example, in a venue where the main reinforcement channel and the supplementary sound channel are working at the same time, the smart sensor located at the boundary between the two is not directly assigned to a certain partition according to its installation position, but is assigned according to the arrival time difference and frequency band direction of the sensor at the acquisition time, thereby reducing misclassification caused by fixed partitions.

[0063] S1.3. Summarize the sound field data within each zone according to the acquisition time, average the sound pressure level of each frequency band, and average the arrival time of the direct sound to obtain the sound field result corresponding to each zone. When the corresponding home channel of the same smart sensor is inconsistent in three adjacent acquisition times, retain the channel identifier that appears most frequently as the home channel of the smart sensor, and re-divide the zones according to the retained home channel and update the sound field result corresponding to each zone. In specific implementation, at each acquisition time, summarize the sound field data of all smart sensors belonging to the same zone, average the sound pressure level of the same frequency band to obtain the average sound pressure level of each frequency band of the zone at that acquisition time; average the arrival time of the direct sound of all smart sensors in the zone to obtain the average sound pressure level of the zone at that acquisition time. The average direct sound arrival time at each acquisition moment forms the sound field result for that zone. Then, to avoid channel jumps caused by short-term reflections, personnel obstruction, or transient interference from individual smart sensors, the channel of the same smart sensor is checked at three adjacent acquisition moments: if the channel is consistent across the three acquisition moments, it is directly retained; if inconsistent, the frequency of each channel identifier is counted across the three acquisition moments, and the channel identifier with the most occurrences is retained as the smart sensor's channel; if the frequency is the same, the channel identifier corresponding to the current acquisition moment is retained. After retention, the zones are re-divided according to the updated channel, and the sound pressure level and direct sound arrival time of each frequency band within each zone are re-summarized to obtain the updated sound field results for each zone. This process suppresses the impact of single misjudgments on the zone results, making the sound field results used in subsequent steps more stable.

[0064] Through the above implementation process, the correspondence between sound field data and tuning parameters can be completed at the same acquisition time. Then, based on the arrival time difference and the consistency of frequency band change direction, the assigned channel of the smart sensor is determined, thereby forming a partition consistent with the actual sound reinforcement response, and outputting the sound field results corresponding to each partition. This preserves real-time response information and reduces the impact of single-point sampling and short-term disturbances on subsequent judgments, thus providing stable input for subsequent change calculation, change source judgment, and parameter adjustment. In practical applications: in conference halls, multi-functional halls, or small performance venues, multiple smart sensors can be deployed in the audience area, near the side walls, and near the stage. The system first reads the sound pressure level and direct current of each smart sensor at the same acquisition time for each frequency band. The system measures the arrival time of sound and reads the gain, equalization, and delay parameters for the main amplification channel, supplementary channel, and monitor channel. Then, it calculates the arrival time difference and the number of frequency bands with the same direction for each channel of each smart sensor, and determines the assigned channel accordingly. Smart sensors with the same assigned channel are grouped into the same partition, and the data within the same partition are averaged to form the sound field result for that partition. When the assigned channel of a smart sensor changes back and forth in three consecutive acquisition times, the channel identifier that appears most frequently is retained, and the partition result is updated accordingly. In this way, the system can generate partitioned sound field results that can be directly used for subsequent automatic sound tuning without relying on manual delineation of fixed measurement areas.

[0065] S2. Compare the sound field results corresponding to each zone with the control results. The control results are the sound field results of the corresponding zone after the tuning is completed. Calculate the change amount of each zone, and determine the source of change based on the change amount during the continuous sampling period. Output the change record corresponding to each zone.

[0066] In this implementation process, step S2 is based on the sound field results already formed in each zone, compared with the control results retained after tuning, to calculate the offset of the current sound field relative to the tuned state, and further determine whether the offset is caused by changes in sound absorption conditions, propagation paths, or transient interference. Specifically, the directly quantifiable change is first calculated by sub-zone and frequency band difference calculations. Then, the continuity and direction of change are analyzed according to the chronological order of continuous sampling periods. Finally, the source of change, along with the corresponding sub-zone, sampling period, and change amount, is written into the change record for subsequent steps to perform classification, storage, and adjustment judgments. This implementation process includes the following steps:

[0067] S2.1. Read the sound field results corresponding to each zone and the corresponding reference results for each zone. Subtract the corresponding sound pressure levels of each frequency band in the reference results from the sound pressure levels of each frequency band in the sound field results, and subtract the corresponding direct sound arrival times in the reference results from the direct sound arrival times in the sound field results. This yields the sound pressure level difference and arrival time difference for each zone in each frequency band, which are used as the corresponding change values ​​for each zone. In practice, first, read the sound field results corresponding to each zone in the current acquisition period according to the zone identifier, and simultaneously read the reference results formed and retained after the tuning of the same zone. Then, perform subtraction operations one by one according to the same frequency band. If a zone contains multiple frequency bands, calculate the sound pressure level difference for each frequency band separately. This process generates a frequency band difference sequence for the current acquisition period. Then, the arrival time of the direct sound in the current sound field result is subtracted from the arrival time of the direct sound in the control result to obtain the arrival time difference for that region. After this processing, each region generates a set of changes that can be directly used for subsequent judgment in each acquisition period. These changes include at least the sound pressure level difference for each frequency band and an arrival time difference. In practical use, if the high-frequency sound pressure level of a certain region decreases relative to the control result, while the low-frequency sound pressure level remains essentially unchanged, and the arrival time of the direct sound does not change, then that region will generate a change result with some frequency band differences changing and the arrival time difference remaining unchanged, for further judgment of the source of the change.

[0068] S2.2. Arrange the changes in the same zone during continuous sampling periods according to the sampling time sequence, and compare the signs of the sound pressure level difference and arrival time difference between adjacent sampling periods for each period. When the sound pressure level difference signs are continuously the same and the arrival time difference remains unchanged, the source of change is determined to be sound absorption change; when the sound pressure level difference signs are continuously the same and the arrival time difference changes continuously in the same direction, the source of change is determined to be propagation change; when the sound pressure level difference signs change or the arrival time difference changes, the source of change is determined to be transient interference, and the corresponding source of change for each zone is output. In specific execution, first, the same... The changes obtained from multiple consecutive sampling periods in a single zone are arranged from front to back according to the sampling time. Then, the previous sampling period is compared with the next sampling period one by one. When comparing the sign of the sound pressure level difference, the difference is determined as positive, negative, or zero for each frequency band. Then, the signs of the main changing frequency bands in two adjacent sampling periods are kept consistent. When comparing the sign of the arrival time difference, it is determined whether the current sampling period remains the same, increases in the same direction, decreases in the same direction, or reverses direction compared to the previous sampling period. When the sign of the sound pressure level difference in the main changing frequency bands remains consistent within the continuous sampling period, and the arrival time difference is always... When the sound pressure level difference remains unchanged, it indicates that the direction of sound energy attenuation continues while the propagation time relationship remains unchanged. In this case, the source of change is determined to be a change in sound absorption. When the sign of the sound pressure level difference in the main frequency bands remains consistent, and the arrival time difference continuously increases or decreases within the continuous sampling period, it indicates that the sound pressure level change and the propagation time change occur synchronously. In this case, the source of change is determined to be a change in propagation. When the sign of the sound pressure level difference in any main frequency band changes from positive to negative, from negative to positive, or from non-zero to zero, or from zero to non-zero, between two consecutive sampling periods, or when the arrival time difference changes from increasing to decreasing, or from decreasing to increasing, the source of change is determined to be a change in propagation. When a change from constant to changing indicates a lack of continuity in the current change, the source of the change is determined to be transient interference. For example, during the gradual entry of the audience, the high-frequency sound pressure level difference in a certain section may be continuously negative while the arrival time difference remains constant, which can be identified as a change in sound absorption. Another example is when a temporary obstruction is added in front of a certain section, and the sound pressure level difference in multiple consecutive sampling periods is continuously negative while the arrival time difference continues to increase, which can be identified as a change in propagation. When someone briefly passes in front of the smart sensor, causing a single obstruction, and then leaves, the direction of the difference between the sampling periods is often reversed, which can be identified as transient interference.

[0069] S2.3. Write the partition identifier, acquisition time period, sound pressure level difference, arrival time difference, and change source of each partition into the corresponding change record. When the change source of the same partition is the same in two adjacent sampling time periods, write a continuous marker into the later change record. When the change source of the same partition is different in two adjacent sampling time periods, write a switching marker into the later change record and output the change record corresponding to each partition. In specific execution, first use the partition identifier + acquisition time period as the positioning field of a change record, and then write the sound pressure level difference, arrival time difference, and the determined change source of each frequency band corresponding to the current acquisition time period into the change record. After completing the writing of the current sampling time period, continue to read the corresponding change source of the previous sampling time period of the same partition. The system records changes and compares the source of change in the previous sampling period with the source of change in the current sampling period. If they are the same, it means that the change state of the partition continues between adjacent sampling periods, and a continuity marker is written into the next change record. If they are different, it means that the change state of the partition changes between adjacent sampling periods, and a switching marker is written into the next change record. After writing all partitions and all sampling periods, the system obtains change records that can be used for subsequent classification, storage, and continuity judgment. The change records formed in this way not only retain the immediate results of each sampling period but also retain the continuity relationship between adjacent sampling periods, which facilitates the direct identification of continuous changes and abrupt changes.

[0070] Through the above implementation process, the current sound field of a zone can be quantified relative to the control state after tuning. Then, based on the continuity of changes within a continuous sampling period, sound absorption changes, propagation changes, and transient interference can be distinguished, and the judgment results can be written into the change record in a structured manner. This avoids misjudgment caused by directly judging based on a single sampling period and provides a continuous and traceable data foundation for subsequent classification and storage by change source, generation of adjustment ranges, and output of adjustment commands. In practical applications: when audiences gradually enter the conference hall, some zones will experience a continuous decrease in mid-to-high frequency sound pressure levels while the arrival time difference remains unchanged. The system will determine the change source corresponding to that zone as a sound absorption change and will then adjust the adjacent zones accordingly. The system continuously records the same source of change and consecutive markers in the sampling period change record. When a temporary screen is added in front of the stage or an equipment vehicle passes in front of the main expansion channel, the sound pressure level difference in the relevant zone will continuously shift and the arrival time difference will continuously change in the same direction. The system will determine the corresponding source of change as a propagation change and write it into the change record. When a staff member passes near a smart sensor for a short time, causing a one-time measurement fluctuation, the direction of the difference between the two sampling periods will change. The system will determine the source of change as a transient interference and write a switching marker in the next change record. Through this processing method, subsequent steps can directly adjust for continuous and effective changes, while maintaining the identification and isolation of short-term disturbances.

[0071] S3. Classify and store the change records corresponding to each partition according to the source of change, and record the partition, time period, direction of change and adjustment range corresponding to each change record, and output the updated record set;

[0072] In this implementation process, step S3 involves organizing and supplementing the change records output in step S2, ensuring that each change record not only retains the source of change and the difference result, but also includes the change direction and adjustment range directly used in subsequent steps. This step mainly completes record classification, sequential arrangement, and field supplementation, without changing the change source already determined in the previous step, but only organizing the change records into a set of records that facilitates subsequent continuous judgment and adjustment instruction generation. This implementation process includes the following steps:

[0073] S3.1 Read the partition identifier, acquisition time period, sound pressure level difference, arrival time difference, and change source from the change records corresponding to each partition. Classify and write each change record according to the change source, partition identifier, and acquisition time period. Arrange the change records with the same partition identifier under the same change source in the order of acquisition time period to obtain a record set. In specific execution, first read the change records output in step S2 one by one, then use the change source as the classification field, the partition identifier as the grouping field, and the acquisition time period as the order field to write the change records to the corresponding positions. For change records with the same change source and the same partition identifier, arrange them from front to back according to the acquisition time period to form an ordered record sequence under the same change source and the same partition. After the above processing, a record set classified by change source and partition identifier and arranged in the order of acquisition time period is obtained.

[0074] S3.2. For each change record in the record set, determine the positive or negative direction of the sound pressure level difference value for each frequency band. Write the positive or negative direction that occurs most frequently as the direction of change, and write the frequency band with a non-zero sound pressure level difference value as the adjustment range. Write the corresponding partition identifier, acquisition time period, direction of change, and adjustment range into each change record, and output the updated record set. In specific execution, read the sound pressure level difference value for each change record for each frequency band. If it is greater than zero, it is recorded as an increasing direction; if it is less than zero, it is recorded as a decreasing direction; if it is equal to zero, it is recorded as no change. Then, count the number of frequency bands that occur in the increasing and decreasing directions, and write the direction that occurs most frequently as the direction of change. Then, write the frequency band with a non-zero sound pressure level difference value as the adjustment range, and write the partition identifier, acquisition time period, direction of change, and adjustment range into the corresponding change record to form the updated record set.

[0075] Through the above implementation process, change records can be organized into a set of records arranged in order by change source, partition identifier, and collection period. The change direction and adjustment range are added to each change record, allowing subsequent steps to directly perform continuity judgment and adjustment processing without re-analyzing the original difference values. This simplifies the subsequent processing and improves the consistency of record usage. In practical applications: when a partition is determined to have the same change source in multiple consecutive sampling periods, the system first writes these change records into the same record sequence. Then, it determines the change direction of the record based on the positive or negative direction of the sound pressure level difference in each frequency band, and writes the frequency band where the change occurred as the adjustment range. In this way, subsequent steps can directly identify continuous changes and generate corresponding adjustment commands based on the change direction and adjustment range in the record set.

[0076] S4. Perform continuity and consistency checks on the change records belonging to the same change source in the updated record set, determine the target record that meets the adjustment conditions, and output the adjustment instruction.

[0077] In this implementation process, step S4 identifies valid changes with continuity and consistency from the record set formed in step S3, and transforms these valid changes into adjustment instructions that can be directly used for subsequent parameter modifications. The processing logic is as follows: first, change records are continuously connected according to partition and time sequence under the same change source; then, continuous records with sufficient duration are selected as target records; finally, the partition identifier, change direction, adjustment range, and corresponding acquisition period from the target records are written into the adjustment instructions. Based on this, the change records of subsequent acquisition periods are further tracked, and the output adjustment instructions are updated or stopped to ensure that the adjustment instructions always correspond to currently ongoing valid changes. This implementation process includes the following steps:

[0078] S4.1 Read all change records from the same source in the record set and arrange them according to the partition identifier and collection time period. Compare the change direction and adjustment range of adjacent change records within the same partition. If the change direction and adjustment range are the same, connect the corresponding change records sequentially as a continuous record; otherwise, end the current continuous record and start connecting again. In specific execution, first read the change records in the record set according to the change source, then group them according to the partition identifier within the same change source, and arrange the change records in each group from front to back according to the collection time period. Then, start the current continuous record from the first change record in each group. Next, read the next change record in sequence and compare the change direction and adjustment range of the next change record with the change direction and adjustment range of the last change record in the current continuous record. If they are the same, it means that the partition maintains the same change trend and the same adjustment object in adjacent collection periods. At this time, connect the next change record to the end of the current continuous record. If they are different, it means that the current continuous relationship ends. At this time, the current continuous record that has been formed is retained, and the connection is restarted with the next change record as the new starting record. After the above processing, one or more continuous records will be formed under the same change source and the same partition, arranged continuously according to the collection period.

[0079] S4.2 Count the number of collection periods contained in each continuous record, and determine the continuous records with no less than two collection periods as target records. Write the partition identifier, change direction, and adjustment range of the target record into the adjustment instruction and output the adjustment instruction. In specific execution, count the number of change records contained in each continuous record formed in S4.1. Since each change record corresponds to one collection period, the number of change records contained in the continuous record is the number of collection periods corresponding to that continuous record. When the number of collection periods reaches two or more, it means that the change is not a single fluctuation, but continues for at least two adjacent collection periods. At this time, the continuous record is determined as the target record. When the number of collection periods is only one, it means that the change only occurs in a single collection period. At this time, it is not included in the adjustment instruction generation process as a target record. After determining the target record, extract the partition identifier, change direction, and adjustment range from it, and write these fields into the adjustment instruction to form the basic adjustment instruction that can be called by subsequent steps. After this processing, subsequent parameter modifications will only be made for continuous changes, and will not directly adjust for single short-term fluctuations.

[0080] S4.3. For continuous records already formed under the same source of change, continue reading the change records corresponding to the next collection period, and compare the change direction and adjustment range of the next change record with the last change record of the continuous record; if the change direction or adjustment range is different, write the continuous record into the target record set, and write the next change record into the first record of the new continuous record; otherwise, continue writing the next change record into the current continuous record; in specific execution, after completing the initial continuous record division, for the same source of change and the same partition where there are still change records for subsequent collection periods, continue reading the next change record in the order of collection periods; compare the next change record with the current continuous record. When comparing the last change record in the continuation record, only the direction of change and the adjustment range are compared, because these two fields directly determine the direction and object of subsequent parameter modifications. If they are the same, it means that the current change is still ongoing, and the next change record is written into the current continuous record. If they are different, it means that the current continuous record should end. In this case, the current continuous record that has already been formed is first written into the target record set, and then the next change record is used as the first record of the new continuous record so that the new continuity judgment can be performed. In this way, the target record set can be continuously maintained as continuous sampling continues, without having to rebuild all continuous records from scratch each time.

[0081] S4.4 Generate adjustment instructions for each target record in the target record set, and write the partition identifier, change source, change direction, adjustment range, and corresponding collection period into the adjustment instructions; when there are two target records with different change sources in the same partition and within the same collection period, retain the adjustment instruction corresponding to the target record with more collection periods and delete the other adjustment instruction; during specific execution, first read the target records in the target record set one by one, and extract the partition identifier, change source, change direction, adjustment range, and collection period range covered by the target record, and write the above fields into the adjustment instructions; among them, the collection period can be written as the start collection period and the end collection period, or it can be written in sequence as all collection periods covered by the target record, as long as it can reflect the target record The duration of the recording is sufficient. After generating all adjustment instructions, continue to check if there are two target records with different sources of change in the same partition during the same sampling period. If so, it means that there are two competing adjustment bases in the partition during the sampling period. At this time, compare the number of sampling periods contained in each of the two target records, retain the adjustment instruction corresponding to the target record with more sampling periods, and delete the other adjustment instruction. The reason for this is that the target record with more sampling periods means that its corresponding change lasts longer and is more stable as an adjustment base. For example, if the same partition continuously shows sound absorption change in multiple adjacent sampling periods, but a single sampling period is determined to be a propagation change, the adjustment instruction corresponding to the sound absorption change will be retained, while the adjustment instruction corresponding to the propagation change will be deleted.

[0082] S4.5 Read the target record corresponding to the output adjustment command, and continue to read the change record of the corresponding partition in the subsequent acquisition period; when the change direction and adjustment range of the subsequent change record are the same as the target record, append the subsequent change record to the target record and update the acquisition period in the corresponding adjustment command; in specific execution, after the adjustment command is output, do not directly regard the adjustment command as fixed, but continue to read the change record generated by the partition in the subsequent acquisition period, and use the target record corresponding to the output adjustment command as the comparison benchmark; when the change direction and adjustment range of the subsequent change record are consistent with the target record, it means that the change corresponding to the target record is still continuing after the adjustment command is output. At this time, append the subsequent change record to the target record and update the acquisition period content in the adjustment command synchronously, so that the adjustment command can reflect the latest continuous interval; after this processing, the adjustment command is not just a one-time judgment result, but can be continuously updated as the subsequent change record continues, thereby avoiding the frequent generation of duplicate commands;

[0083] S4.6 When the direction of change of a subsequent change record is opposite to that of the target record, or the adjustment range of the subsequent change record is different from that of the target record, the appending and writing stops, and the original adjustment instruction is rewritten as a stop execution instruction. The subsequent change record is written back to the record set as a new judgment starting point. In specific execution, when reading a subsequent change record, if it is found that its direction of change is opposite to that of the current target record, or its adjustment range is different from that of the current target record, it means that the continuous change corresponding to the current target record has ended, or the nature of the change has changed. At this time, the subsequent change record is no longer appended to the original target record. Instead, the execution of the original adjustment instruction is stopped, and the original adjustment instruction is rewritten as a stop execution instruction to notify the subsequent parameter modification steps that the original adjustment basis is no longer used. At the same time, the subsequent change record is written back to the record set and re-participates in the processing of S4.1 to S4.4 as a new judgment starting point. This ensures that the subsequent steps always adjust based on the currently valid continuous change, and do not mistakenly incorporate changes that have reversed or changed in range into the original instruction.

[0084] Through the above implementation process, effective changes with continuity and consistency can be selected from the record set and stably transformed into adjustment instructions. Simultaneously, when subsequent sampling occurs, the adjustment instructions are updated or stopped. This avoids the system responding too quickly to a single disturbance and maintains the continuity of the adjustment basis when changes continue, thereby improving the targeting and stability of subsequent parameter modifications. In practical applications: when a certain partition exhibits the same source of change in multiple consecutive sampling periods, and the direction of change is consistently decreasing with the adjustment range consistently concentrated in the mid-to-high frequency range, the system first connects these change records into continuous records, and then, when the number of sampling periods reaches two or more... The system then identifies the target record and generates a corresponding adjustment instruction. If the partition maintains the same direction of change and adjustment range in subsequent sampling periods, the system continues to append subsequent change records to the target record and synchronously updates the sampling period of the adjustment instruction. If the direction of change of the partition changes to increase in subsequent sampling periods, or the adjustment range changes from medium-high frequency to low frequency, the system stops appending and writing, rewrites the original adjustment instruction as a stop execution instruction, and writes the new change record back to the record set as the starting point for the next round of continuity judgment. Through this processing method, the adjustment instruction output by the system always corresponds to a continuously effective partition change, and will not continue to use an unsuitable adjustment basis due to short-term fluctuations or changes in the nature of the change.

[0085] S5. Based on the adjustment command, determine the target parameters corresponding to the target record from the tuning parameters, modify only the gain value, equalization parameter or delay parameter corresponding to the adjustment range in the corresponding partition of the target record, and output the updated tuning parameters.

[0086] In this implementation process, step S5 is to select the parameter items that need to be modified from the currently effective tuning parameters of the audio system according to the adjustment command output in step S4, and perform corresponding modifications according to the source, direction, and adjustment range of the change, and then write the modification results back to the corresponding partition. The focus of this process is not to adjust all tuning parameters uniformly, but to first determine the partition and parameter type to be processed according to the adjustment command, and then only modify the parameters directly corresponding to the partition, the change, and the adjustment range, thereby ensuring that the parameter modification and the change record correspond one-to-one and avoiding unrelated partitions from being affected. This implementation process includes the following steps:

[0087] S5.1 Read the zone identifier, change source, change direction, and adjustment range from the adjustment command, and read the corresponding channel's gain value, equalization parameter, and delay parameter from the tuning parameters according to the zone identifier; when the change source is sound absorption change, determine the equalization parameter and gain value corresponding to the adjustment range as the target parameters; when the change source is propagation change, determine the delay parameter and gain value corresponding to the adjustment range as the target parameters; in specific execution, first read the adjustment commands output in step S4 one by one, and extract the zone identifier, change source, change direction, and adjustment range from each adjustment command; then, according to the zone identifier, find the channel parameter corresponding to that zone in the current tuning parameters, and read the currently effective gain value, equalization parameter, and delay parameter of that channel; then, execute parameter selection according to the change source: When the source of change is a change in sound absorption, it indicates that the current change mainly manifests as a change in the sound energy distribution within the zone, while the propagation time relationship remains unchanged. In this case, the equalization parameters and gain values ​​of the corresponding frequency band within the adjustment range are determined as the target parameters. When the source of change is a change in propagation, it indicates that the current change is reflected not only in the sound pressure level but also in the arrival time relationship. In this case, the delay parameter and the gain value of the corresponding frequency band within the adjustment range are determined as the target parameters. After the above processing, each adjustment command corresponds to a set of clear target parameters for direct modification in the next step. For example, when the source of change in a certain zone is a change in sound absorption and the adjustment range is mid-frequency and high-frequency, the system will select the mid-frequency and high-frequency equalization parameters of the corresponding channel of that zone and the current gain value as the target parameters, without simultaneously selecting the delay parameter.

[0088] S5.2 Modify the target parameters item by item according to the direction of change; when the direction of change is increasing, decrease the gain value or equalization parameter of the corresponding frequency band within the adjustment range step by step, or decrease the delay parameter in the opposite direction of the arrival time difference; when the direction of change is decreasing, increase the gain value or equalization parameter of the corresponding frequency band within the adjustment range step by step, or increase the delay parameter in the same direction as the arrival time difference; in specific execution, first read the direction of change corresponding to the current adjustment command, and then process the determined target parameters item by item; when the target parameter is a gain value or equalization parameter, modify it band by band according to the adjustment range: if the direction of change is increasing, it means that the sound pressure level of the current partition in the corresponding frequency band is higher than the reference result, at this time, adjust the gain value or equalization parameter of the corresponding frequency band in the decreasing direction according to the current setting value; if the direction of change is decreasing, it means that the sound pressure level of the current partition in the corresponding frequency band is lower than the reference result, at this time, adjust the gain value or equalization parameter of the corresponding frequency band in the increasing direction according to the current setting value. When the target parameters include a delay parameter, it is also necessary to read the corresponding arrival time difference value in the current change record of the partition, and use the direction of the arrival time difference value as the basis for delay adjustment: if the arrival time difference value is positive, it is adjusted according to the rule of decreasing in the opposite direction or increasing in the same direction; if the arrival time difference value is negative, it is adjusted in the opposite direction. Here, step-by-step means that in a single processing process, the parameter is not directly jumped to a new extreme value, but is gradually adjusted based on the current parameter in the manner of one modification corresponding to one write, and the parameter value after each modification is used as the input for the next calculation. This processing can avoid new sound field fluctuations caused by a large change in the parameter at one time. For example, when a partition shows a continuous increase in the mid-high frequency and the source of the change is a change in sound absorption, the system will gradually decrease the mid-high frequency equalization parameter and gain value; when a partition shows a propagation change and the arrival time difference value continues to increase, the system will adjust the gain value corresponding to the adjustment range while decreasing the delay parameter in the opposite direction.

[0089] S5.3 Write the modified target parameters back to the tuning parameters of the corresponding partition, while keeping the tuning parameters of other partitions unchanged, and output the updated tuning parameters. In specific execution, after completing the modification of the target parameters, first locate the storage location of the corresponding channel of the partition in the current tuning parameters according to the partition identifier, and then write the modified gain value, equalization parameter, and delay parameter back to the corresponding channel of the partition one by one. For other partitions not selected by the current adjustment command, do not perform any modification on their tuning parameters, and keep their parameter values ​​consistent with those before the modification. After completing the write-back of the parameters corresponding to the current adjustment command, recombine the tuning parameters of all partitions to form the updated tuning parameters, and use them as input for subsequent verification steps. After this processing, the scope of parameter modification is always limited to the partition corresponding to the current adjustment command, and will not change the parameters of other areas due to the change of a single area, thus ensuring the independence of partition adjustment.

[0090] Through the above implementation process, the adjustment command output in step S4 can be accurately converted into specific parameter modification actions in the current tuning parameters. The gain value, equalization parameter, and delay parameter are selected and modified according to the source and direction of the change, respectively. Simultaneously, the modification results are written back only to the corresponding partition, thus achieving a direct correspondence between parameter modification and partition changes. This ensures the targeted nature of parameter modification while avoiding cascading changes to parameters in unrelated partitions, providing clear input for subsequent result verification based on the updated tuning parameters. In practical applications: when a partition exhibits sound absorption changes over multiple consecutive sampling periods, and the direction of change is downward with the adjustment range concentrated in the mid and high frequencies, the system first... The system reads the gain, equalization, and delay parameters of the corresponding channel for each zone based on its zone identifier. Then, it identifies the equalization parameters for the mid-frequency and high-frequency bands, along with the current gain, as the target parameters and modifies them sequentially in the direction of increase. Finally, it writes the modified mid-frequency and high-frequency parameters and gain values ​​back to the corresponding channel for that zone. Similarly, when a zone is identified as having a propagation change, the system reads the gain value of the corresponding frequency band, as well as the current delay parameter. It then modifies the delay parameter based on the arrival time difference direction and writes the modified parameters back to that zone. Through this processing method, the system can translate the continuous and valid changes identified in the preceding steps into actual tuning parameter modifications only for the corresponding zone.

[0091] S6. Based on the updated tuning parameters and the sound field data subsequently collected by the smart sensor, verify the sound field results after parameter modification, and when the corresponding sound field change disappears or the subsequent sound field data indicates that the original source of change is not valid, cancel or pause the local adjustment corresponding to the target record and output the automatic tuning result.

[0092] In this implementation process, step S6 serves to review the result of parameter modification completed in step S5, determining whether the current parameter modification weakens, eliminates, or alters the nature of the sound field change in the corresponding zone, and accordingly decides whether to retain, pause, or cancel the parameter modification. The processing logic is as follows: first, sound field data is reacquired in subsequent acquisition periods, and the aforementioned zoning and comparison methods are used to obtain new sound field results, changes, and sources of change for each zone. Then, these results are compared with the original changes corresponding to the target record. Finally, the comparison conclusion is converted into parameter status and written into the automatic tuning result. After this processing, the system does not end directly after parameter modification, but forms a closed loop through re-acquisition and re-judgment, thereby ensuring that parameter modification always corresponds to the current effective change. This implementation process includes the following steps:

[0093] S6.1 Read the updated tuning parameters and the sound field data collected by the smart sensors during the subsequent acquisition period. Divide the sound field data into partitions according to the response magnitude of each smart sensor to each channel, output the sound field results corresponding to each partition, and compare the sound field results corresponding to each partition with the corresponding comparison results to obtain the amount of change and the source of change of each partition during the subsequent acquisition period. In specific execution, first read the updated tuning parameters output in step S5, and at the same time read the sound field data collected by the smart sensors during the subsequent acquisition period. Then, using the partitioning method in step S1, re-partition the sound field data during the subsequent acquisition period, that is, determine the channel according to the response magnitude of each smart sensor to each channel, and divide the sound field data belonging to the same channel into the same partition, and then divide each frequency band in each partition. The sound pressure level and direct sound arrival time are summarized to obtain the sound field results of each zone in the subsequent acquisition period. After the zoning is completed, the comparison method in step S2 is used again to compare the sound field results of each zone in the subsequent acquisition period with the corresponding zone's comparison results item by item, calculate the sound pressure level difference and arrival time difference of each frequency band, and determine the current source of change of the zone according to the continuous sampling judgment rule. After this processing, the new amount of change and the source of change of each zone in the subsequent acquisition period after the parameter modification takes effect can be obtained as the direct basis for the next state judgment. For example, if a zone shows that the mid-high frequency sound pressure level is consistently low before the parameter modification, and the mid-high frequency difference is significantly reduced when recalculated in the subsequent acquisition period after the parameter modification is completed, it indicates that the modified tuning parameters have had a corrective effect on the zone.

[0094] S6.2. Compare the changes and sources of change in each partition during subsequent acquisition periods with the changes and sources of change corresponding to the target record. When the change is zero, cancel the parameter modification corresponding to the target record. When the source of change is different from the source of change in the target record, pause the parameter modification corresponding to the target record and retain the parameter value before modification. When the change decreases and the source of change remains unchanged, maintain the parameter modification corresponding to the target record. In specific execution, first read the target record corresponding to the current partition and obtain the original change and original source of change written in the target record, then compare the changes and sources of change obtained in subsequent acquisition periods with it. When the change in subsequent acquisition periods is zero, it means that the current sound field result of the partition is no different from the control result. At this time, it means that the original change has disappeared, and it is unnecessary to continue to retain the parameter modification. Therefore, the parameter modification corresponding to the target record is written to the canceled state. When the source of change in subsequent acquisition periods is different from the source of change in the target record, for example, the original change has disappeared. If a change in sound absorption is subsequently determined to be a propagation change, or if a change in propagation is subsequently determined to be transient interference, it indicates that the nature of the change in the current zone has changed. In this case, the original parameter modifications are no longer used; instead, the corresponding parameter modifications are paused, and the parameter values ​​retained before the modification are restored. When the amount of change in subsequent sampling periods decreases relative to the original amount of change in the target record, and the source of change remains unchanged, it indicates that the original change still exists but its degree has decreased, and the direction of the current parameter modification is still consistent with the nature of the change. Therefore, the parameter modification corresponding to the target record is maintained. Through the above comparison rules, subsequent sampling results can be directly converted into parameter states without regenerating new adjustment instructions. For example, if a zone originally experienced a decrease in mid-to-high frequency sound pressure level due to a change in sound absorption, and the mid-to-high frequency difference narrows after parameter modification and is still determined to be a change in sound absorption, the current modification is maintained. If there is no difference subsequently, the modification is revoked. If the subsequent determination changes to a propagation change, the original modification is paused, and the system waits for the next round of judgment.

[0095] S6.3. Write the parameter status after cancellation, pause, or hold into the tuning parameters of the corresponding partition, and write the corresponding partition identifier, subsequent acquisition period, change amount, change source, and parameter status into the automatic tuning result, and output the automatic tuning result; In specific execution, after completing the parameter status determination in step S6.2, first find the current corresponding tuning parameter storage location of the partition according to the partition identifier, and then perform the write according to the parameter status; When the parameter status is canceled, restore the partition parameter to the parameter value retained before executing step S5; When the parameter status is paused, similarly restore to the parameter value before modification, and write in the status field. The process is paused upon input; if the parameter status is "hold," the modified parameter value is retained without rollback; after writing the parameters, the partition identifier, subsequent acquisition period, current change amount, current change source, and parameter status are written to the automatic tuning result; after performing the same processing on all partitions, the automatic tuning result is output for the system to record the status after the current automatic tuning execution and subsequent calls; after this processing, the automatic tuning result not only indicates whether the current parameter is retained, but also retains the changes and final status of the corresponding partition in the subsequent acquisition period, which is convenient for subsequent tracking or for maintenance personnel to view;

[0096] Through the above implementation process, the sound field results for subsequent acquisition periods can be reacquired after parameter modification. By directly comparing the changes with the corresponding changes in the target record, it can be determined whether the current parameter modification should be revoked, paused, or maintained, thus forming a complete parameter modification closed loop. This avoids parameter modifications being retained for a long time and becoming detached from actual changes, and also allows the current parameters to be maintained while changes continue and modifications are effective, thereby improving the stability and adaptability of automatic tuning results. In practical applications: when the mid-high frequency sound pressure level of a certain zone remains low due to changes in sound absorption, the system modifies the mid-high frequency equalization parameters and gain values ​​of that zone according to step S5. Subsequently, the system will reread the sound field data of the partition and calculate the new changes and sources of change during subsequent acquisition periods. If the difference in mid-to-high frequency sound pressure level has returned to zero at this time, the system will cancel the parameter modification. If the difference still exists but has decreased significantly, and the source of change is still sound absorption change, the system will retain the parameter modification. If subsequent sampling shows that the partition no longer exhibits sound absorption change, but instead becomes propagation change or transient interference, the system will pause the original modification and restore the parameter value before the modification. Through this processing method, the automatic tuning process can dynamically determine the retention status of the parameters based on the subsequent actual sound field feedback, rather than keeping them fixed after completing a parameter modification.

[0097] Furthermore, it also includes an automatic audio tuning system based on smart sensors, including:

[0098] The data acquisition module acquires sound field data collected by various intelligent sensors deployed in the target sound reinforcement venue, as well as the tuning parameters of the audio system.

[0099] The partitioning module partitions the sound field data according to the response magnitude of each smart sensor to each channel and outputs the sound field results corresponding to each partition.

[0100] The change analysis module compares the sound field results corresponding to each zone with the control results, which are the sound field results of the corresponding zone after tuning. It calculates the change amount of each zone, determines the source of change based on the change amount during the continuous sampling period, and outputs the change record corresponding to each zone.

[0101] The record processing module categorizes and stores the change records corresponding to each partition according to the source of change, and records the partition, time period, direction of change and adjustment range corresponding to each change record, and outputs the updated record set;

[0102] The instruction generation module performs continuity and consistency checks on change records belonging to the same source of change in the updated record set, determines the target record that meets the adjustment conditions, and outputs the adjustment instruction.

[0103] The parameter adjustment module determines the target parameters corresponding to the target record from the tuning parameters according to the adjustment command, and only modifies the gain value, equalization parameter or delay parameter corresponding to the adjustment range in the corresponding partition of the target record, and outputs the updated tuning parameters.

[0104] The result verification module verifies the sound field results after parameter modification based on the updated tuning parameters and the sound field data subsequently collected by the smart sensor. When the corresponding sound field change disappears or the subsequent sound field data indicates that the original source of the change is not valid, the module cancels or pauses the local adjustment corresponding to the target record and outputs the automatic tuning result.

[0105] Working Principle: This invention first deploys multiple intelligent sensors within the sound reinforcement venue to collect sound field data from different locations in real time, and simultaneously reads the currently effective tuning parameters of each channel of the audio system. Then, based on the response relationship between each intelligent sensor and each channel, the system divides the collected sound field data into different zones, forming the sound field results corresponding to each zone. After obtaining the zone results, the current zone result is compared with the control result after tuning, calculating the change in each zone, and combining the change trend over multiple consecutive sampling periods to determine whether the change is caused by changes in sound absorption conditions, propagation paths, or transient interference. Next, the system organizes these change records into a continuously traceable record set, filters out continuous and consistent effective changes, and generates corresponding adjustment instructions. Then, based on the adjustment instructions, only the gain value, equalization parameters, or delay parameters of the relevant frequency bands within the corresponding zone are modified. Finally, the system continues to collect subsequent sound field data and re-verifies the modified results. If the change decreases, the current modification is maintained; if the change disappears, the modification is revoked; if the nature of the change has changed, the original modification is paused, thus forming a closed-loop tuning process that can automatically judge, automatically adjust, and automatically check back.

[0106] For example, in conference halls or small performance venues, as the audience gradually enters, the mid-to-high frequency sound often weakens due to sound absorption by the human body. Traditionally, this requires sound engineers to manually compensate while listening. In this solution, intelligent sensors deployed in the audience area, near the side walls, and in the front row first detect a continuous decrease in the mid-to-high frequency sound pressure level relative to the state when the sound was first tuned. Based on this, the system determines that a sound absorption change has occurred in that area and confirms that this change persists for multiple consecutive sampling periods. Subsequently, the system automatically adjusts the mid-to-high frequency equalization parameters and gain values ​​of the corresponding channel for that area, bringing the sound in that area back to the state when the sound was first tuned. If the audience is seated and the sound field returns to balance, the system will find that the change has disappeared in subsequent verifications and will stop retaining any unnecessary compensation. If an obstruction is temporarily added to the scene, causing the change to change from a sound absorption change to a propagation change, the system will recognize this change in the nature of the change, pause the original compensation, and re-enter the next round of judgment. Therefore, in practical use, this solution acts like an automatic sound tuning assistant that continuously observes the scene, determines the cause, makes local corrections, and reconfirms, replacing a large amount of manual and repetitive adjustment work in dynamic environments.

[0107] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An automatic audio tuning method based on intelligent sensors, characterized in that, include: S1. Acquire the sound field data collected by each intelligent sensor deployed in the target sound reinforcement site and the tuning parameters of the audio system. Divide the sound field data into partitions according to the response magnitude of each intelligent sensor to each channel, and output the sound field results corresponding to each partition. S2. Compare the sound field results corresponding to each zone with the control results. The control results are the sound field results of the corresponding zone after the tuning is completed. Calculate the change amount of each zone, and determine the source of change based on the change amount during the continuous sampling period. Output the change record corresponding to each zone. S3. Classify and store the change records corresponding to each partition according to the source of change, and record the partition, time period, direction of change and adjustment range corresponding to each change record, and output the updated record set; S4. Perform continuity and consistency checks on the change records belonging to the same change source in the updated record set, determine the target record that meets the adjustment conditions, and output the adjustment instruction. S5. Based on the adjustment command, determine the target parameters corresponding to the target record from the tuning parameters, modify only the gain value, equalization parameter or delay parameter corresponding to the adjustment range in the corresponding partition of the target record, and output the updated tuning parameters. S6. Based on the updated tuning parameters and the sound field data subsequently collected by the smart sensor, verify the sound field results after parameter modification, and when the corresponding sound field change disappears or the subsequent sound field data indicates that the original source of change is not valid, cancel or pause the local adjustment corresponding to the target record, and output the automatic tuning result.

2. The method according to claim 1, characterized in that, S1 includes: S1.1 Read the installation location, acquisition time, sound pressure level of each frequency band and arrival time of direct sound collected by each smart sensor as sound field data; read the channel identifier, gain value, equalization parameter and delay parameter of each channel of the audio system as tuning parameters; output the sound field data and tuning parameters aligned with the acquisition time. S1.2 Calculate the response size for each channel of each smart sensor. The response size is determined according to the following rules: Subtract the delay parameter of the corresponding channel from the arrival time of the direct sound of the smart sensor to obtain the arrival time difference. Compare the direction of change of the sound pressure level of each frequency band of the smart sensor with the direction of change of the gain value and equalization parameter of the corresponding channel band by band to obtain the number of frequency bands with the same direction. First sort them by the absolute value of the arrival time difference from smallest to largest. Then, when the absolute values ​​of the arrival time differences are the same, sort them by the number of frequency bands with the same direction from most to least. Identify the channel with the first ranking as the channel to which the smart sensor belongs, and divide the sound field data with the same channel into the same partition. S1.

3. Summarize the sound field data in each zone according to the acquisition time, take the average value of the sound pressure level of each frequency band and the average value of the arrival time of the direct sound to obtain the sound field results corresponding to each zone; when the corresponding home channel of the same smart sensor is inconsistent in three adjacent acquisition times, retain the channel identifier that appears most frequently as the home channel of the smart sensor, and re-divide the zones according to the retained home channel and update the sound field results corresponding to each zone.

3. The method according to claim 2, characterized in that, S2 includes: S2.1 Read the sound field results corresponding to each partition and the corresponding comparison results of the partition. Subtract the sound pressure level of each frequency band in the comparison results from the sound pressure level of each frequency band in the sound field results, and subtract the arrival time of the direct sound in the comparison results from the arrival time of the direct sound in the sound field results to obtain the sound pressure level difference and arrival time difference of each partition in each frequency band, which are used as the change amount corresponding to the partition. S2.

2. Arrange the changes in the same zone during continuous sampling periods in the order of sampling time, and compare the signs of the sound pressure level difference and the arrival time difference between two adjacent sampling periods for each period. When the signs of the sound pressure level difference are continuously the same and the arrival time difference remains unchanged, the source of change is determined to be sound absorption change. When the signs of the sound pressure level difference are continuously the same and the arrival time difference changes continuously in the same direction, the source of change is determined to be propagation change. When the signs of the sound pressure level difference change before and after or the arrival time difference changes before and after, the source of change is determined to be transient interference, and the corresponding source of change for each zone is output. S2.3 Write the partition identifier, acquisition time period, sound pressure level difference, arrival time difference, and change source into the corresponding change record; when the change source of the same partition is the same in two adjacent sampling time periods, write a continuous mark in the next change record; when the change source of the same partition is different in two adjacent sampling time periods, write a switching mark in the next change record, and output the change record corresponding to each partition.

4. The method according to claim 3, characterized in that, The step S3 includes: S3.1 Read the partition identifier, collection period, sound pressure level difference, arrival time difference and change source from the change records corresponding to each partition. Classify and write each change record according to the change source, partition identifier and collection period. Arrange the change records with the same partition identifier under the same change source in the order of collection period to obtain the record set. S3.2 Determine the positive and negative directions of the sound pressure level difference for each change record in the record set for each frequency band. Write the positive and negative directions that occur most frequently as the direction of change, write the frequency bands where the sound pressure level difference is not zero as the adjustment range, and write the corresponding partition identifier, acquisition time period, direction of change and adjustment range into each change record. Output the updated record set.

5. The method according to claim 4, characterized in that, S4 includes: S4.1 Read all change records with the same change source in the record set and arrange them in order of partition identifier and collection time period; compare the change direction and adjustment range of adjacent change records in the same partition. When the change direction and adjustment range are the same, connect the corresponding change records into continuous records in sequence. Otherwise, end the current continuous record and start connecting again. S4.2 Count the number of collection periods included in each continuous record, and determine the continuous records with no less than two collection periods as target records. Write the partition identifier, change direction and adjustment range of the target records into the adjustment command and output the adjustment command.

6. The method according to claim 5, characterized in that, S4 further includes: S4.

3. For continuous records formed under the same source of change, continue to read the change records corresponding to the next collection period, and compare the change direction and adjustment range of the next change record with the last change record of the continuous record; when the change direction or adjustment range is different, write the continuous record into the target record set, and write the next change record into the first record of the new continuous record; otherwise, continue to write the next change record into the current continuous record. S4.4 Generate adjustment instructions for each target record in the target record set, and write the partition identifier, change source, change direction, adjustment range and corresponding collection time period into the adjustment instructions; when there are two target records with different change sources in the same partition in the same collection time period, retain the adjustment instructions corresponding to the target record with more collection time periods and delete the other adjustment instructions.

7. The method according to claim 6, characterized in that, S4 further includes: S4.5 Read the target record corresponding to the output adjustment command, and continue to read the change record of the corresponding partition in the subsequent acquisition period; when the change direction and adjustment range of the subsequent change record are the same as the target record, append the subsequent change record to the target record and update the acquisition period in the corresponding adjustment command; S4.6 When the direction of change of subsequent change records is opposite to that of the target record, or the adjustment range of subsequent change records is different from that of the target record, the append writing stops, the original adjustment instruction is written as a stop execution instruction, and the subsequent change records are written back to the record set as the new judgment starting point.

8. The method according to claim 7, characterized in that, S5 includes: S5.1 Read the zone identifier, change source, change direction, and adjustment range from the adjustment command, and read the gain value, equalization parameter, and delay parameter of the corresponding channel from the tuning parameters according to the zone identifier; when the change source is sound absorption change, determine the equalization parameter and gain value corresponding to the adjustment range as the target parameter; when the change source is propagation change, determine the delay parameter and gain value corresponding to the adjustment range as the target parameter. S5.2 Modify the target parameters one by one according to the direction of change; when the direction of change is increasing, decrease the gain value or equalization parameter of the corresponding frequency band within the adjustment range step by step, or decrease the delay parameter in the opposite direction of the arrival time difference; when the direction of change is decreasing, increase the gain value or equalization parameter of the corresponding frequency band within the adjustment range step by step, or increase the delay parameter in the same direction as the arrival time difference. S5.3 Write the modified target parameters back to the tuning parameters of the corresponding partition, while keeping the tuning parameters of other partitions unchanged, and output the updated tuning parameters.

9. The method according to claim 8, characterized in that, S6 includes: S6.1 Read the updated tuning parameters and the sound field data collected by the smart sensor in the subsequent acquisition period. Divide the sound field data into partitions according to the response of each smart sensor to each channel. Output the sound field results corresponding to each partition. Compare the sound field results corresponding to each partition with the corresponding partition's comparison results to obtain the amount of change and the source of change of each partition in the subsequent acquisition period. S6.

2. Compare the amount of change and the source of change of each partition in the subsequent collection period with the amount of change and the source of change corresponding to the target record; when the amount of change is zero, cancel the parameter modification corresponding to the target record; when the source of change is different from the source of change in the target record, pause the parameter modification corresponding to the target record and retain the parameter value before modification; when the amount of change decreases and the source of change remains unchanged, maintain the parameter modification corresponding to the target record. S6.3 Write the parameter status after cancellation, pause or hold into the tuning parameters of the corresponding partition, and write the corresponding partition identifier, subsequent acquisition period, change amount, change source and parameter status into the automatic tuning result, and output the automatic tuning result.

10. An automatic audio tuning system based on intelligent sensors, characterized in that, To implement the method according to any one of claims 1-9, comprising: The data acquisition module acquires sound field data collected by various intelligent sensors deployed in the target sound reinforcement venue, as well as the tuning parameters of the audio system. The partitioning module partitions the sound field data according to the response magnitude of each smart sensor to each channel and outputs the sound field results corresponding to each partition. The change analysis module compares the sound field results corresponding to each zone with the control results, which are the sound field results of the corresponding zone after tuning. It calculates the change amount of each zone, determines the source of change based on the change amount during the continuous sampling period, and outputs the change record corresponding to each zone. The record processing module categorizes and stores the change records corresponding to each partition according to the source of change, and records the partition, time period, direction of change and adjustment range corresponding to each change record, and outputs the updated record set; The instruction generation module performs continuity and consistency checks on change records belonging to the same source of change in the updated record set, determines the target record that meets the adjustment conditions, and outputs the adjustment instruction. The parameter adjustment module determines the target parameters corresponding to the target record from the tuning parameters according to the adjustment command, and only modifies the gain value, equalization parameter or delay parameter corresponding to the adjustment range in the corresponding partition of the target record, and outputs the updated tuning parameters. The result verification module verifies the sound field results after parameter modification based on the updated tuning parameters and the sound field data subsequently collected by the smart sensor. When the corresponding sound field change disappears or the subsequent sound field data indicates that the original source of the change is not valid, the module cancels or pauses the local adjustment corresponding to the target record and outputs the automatic tuning result.

Images