Building curtain wall sound insulation performance detection system and method
通过人工智能分析建筑幕墙噪声源和隔音后声音强度的时序关联,实现了建筑幕墙隔音性能的智能化自动检测,解决了传统检测方法的主观性和效率低的问题,提高了检测准确性和效率 .
Patent Information
- Application Number
- CN202410524463.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-29
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-04-29
AI Technical Summary
Traditional architectural curtain wall sound insulation performance detection methods require manual participation, which are cumbersome and time-consuming, easily affected by subjective factors, and it is difficult to comprehensively evaluate the sound insulation effect under different noise conditions.
The first acoustic device monitors the noise intensity of the noise source in real time, the second acoustic device monitors the sound intensity after sound insulation by the building curtain wall in real time, and introduces data processing and analysis algorithms based on artificial intelligence on the back end to perform the noise intensity of the noise source in real time, so as to carry out the noise intensity of the noise source in real time, and introduces artificial intelligence-based data processing and analysis algorithms on the back end to perform the noise intensity of the noise source. The timing coordination and responsive correlation analysis is conducted between sound intensity to determine the noise reduction performance level of building curtain walls.
It realizes intelligent automatic detection of sound insulation performance of building curtain walls, avoids manual errors and inefficiency problems, and can test noise reduction performance under different noise timing conditions, improving detection efficiency and accuracy.
Smart Images

Figure CN118330038B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of sound insulation performance detection, and in particular to a system and method for detecting sound insulation performance of a building curtain wall. Background Art
[0002] In construction projects, curtain walls are decorative, heat-insulating, heat-insulating and waterproof components of building exterior walls, usually made of glass, metal, stone, ceramics and other materials. Curtain walls are not only an important part of the building's exterior, but also have the functions of protecting the building's interior space, regulating indoor and outdoor temperatures, and isolating noise.
[0003] With the continuous acceleration of urbanization and the increase in building density, noise pollution in cities is becoming increasingly serious. The sound insulation performance of building curtain walls is crucial to ensuring the comfort and quiet environment inside the building. Therefore, it is of great significance to conduct a scientific and accurate evaluation of the sound insulation performance of building curtain walls. However, traditional building curtain wall sound insulation performance testing schemes usually require manual participation, the measurement process is cumbersome and time-consuming, and is easily affected by subjective factors, and the results may not be objective and accurate enough. In addition, traditional building curtain wall sound insulation performance testing methods can usually only provide static sound insulation performance data, which makes it difficult to fully evaluate the actual sound insulation effect of building curtain walls under different noise conditions.
[0004] Therefore, an optimized building curtain wall sound insulation performance detection system is desired. Summary of the invention
[0005] The present disclosure is made in view of the above problems. One object of the present disclosure is to provide a system and method for detecting the sound insulation performance of a building curtain wall.
[0006] An embodiment of the present disclosure provides a building curtain wall sound insulation performance detection system, which includes:
[0007] A noise source noise intensity data acquisition module, used to acquire a time series of noise intensity of the noise source collected by the first acoustic device;
[0008] A noise intensity data acquisition module after sound insulation, used to obtain the time series of the sound intensity after sound insulation of the building curtain wall collected by the second acoustic device;
[0009] A noise intensity data time series arrangement module, used for arranging the time series of the noise intensity and the time series of the sound intensity into a noise intensity time series input vector and a sound intensity time series input vector respectively according to the time dimension;
[0010] A noise time series vector segmentation module, used for performing vector segmentation on the noise intensity time series input vector and the sound intensity time series input vector at the same time scale to obtain a sequence of noise intensity local time series input vectors and a sequence of sound intensity local time series input vectors;
[0011] A noise reduction effect local time series response feature extraction module, used for inputting each group of corresponding noise intensity local time series input vectors and sound intensity local time series input vectors in the sequence of the noise intensity local time series input vectors and the sequence of the sound intensity local time series input vectors into a gamma correction response association module to obtain a sequence of noise reduction effect local time series response feature vectors;
[0012] A noise reduction effect full-time fusion feature characterization module, used for performing adaptive attention weight allocation fusion processing on the sequence of the noise reduction effect local time series response feature vectors to obtain a noise reduction effect full-time fusion feature; and
[0013] The noise reduction performance level detection module is used to determine the noise reduction performance level label based on the full temporal fusion characteristics of the noise reduction effect.
[0014] For example, according to the building curtain wall sound insulation performance detection system of the embodiment of the present disclosure, the noise reduction effect local time series response feature extraction module is used to:
[0015] Each group of corresponding noise intensity local time series input vectors and sound intensity local time series input vectors in the sequence of the noise intensity local time series input vectors and the sequence of the sound intensity local time series input vectors is input into the gamma correction response association module for processing using the following gamma correction response association formula to obtain the sequence of the noise reduction effect local time series response feature vectors; wherein the gamma correction response association formula is:
[0016]
[0017] in, and The first sequence of each corresponding noise intensity local time series input vector and sound intensity local time series input vector in the sequence of the noise intensity local time series input vector and the sequence of the sound intensity local time series input vector are respectively The feature values at each position, A, B, C and D are the hyperparameters for adjustment. is the first of the local time series response feature vectors of the noise reduction effect in the sequence of the local time series response feature vectors of the noise reduction effect. The feature value at each position.
[0018] For example, according to the building curtain wall sound insulation performance detection system of the embodiment of the present disclosure, the noise reduction effect full time series fusion feature characterization module is used to:
[0019] The sequence of the local temporal response feature vectors of the noise reduction effect is passed through an adaptive attention weight fusion network to obtain a full temporal fusion feature vector of the noise reduction effect as the full temporal fusion feature of the noise reduction effect.
[0020] For example, according to the building curtain wall sound insulation performance detection system of the embodiment of the present disclosure, the noise reduction effect full time series fusion feature characterization module is used to:
[0021] The sequence of the local temporal response feature vectors of the noise reduction effect is processed by the adaptive attention weight fusion network using the following adaptive attention fusion formula to obtain the full temporal fusion feature vector of the noise reduction effect; wherein the adaptive attention fusion formula is:
[0022]
[0023]
[0024]
[0025] in, is the first in the sequence of local temporal response feature vectors of the noise reduction effect The local time series response feature vector of the noise reduction effect, and Represent the weight coefficient vector and weight coefficient matrix respectively, is the offset vector, express function, For the The attention score of the local temporal response feature vector of the noise reduction effect, and are all hyperparameters, express function, is the number of vectors in the sequence of the local temporal response feature vectors of the noise reduction effect, It is the full temporal fusion feature vector of the noise reduction effect.
[0026] For example, according to the building curtain wall sound insulation performance detection system of the embodiment of the present disclosure, the noise reduction performance level detection module is used to:
[0027] The full-time fusion feature vector of the noise reduction effect is passed through a noise reduction performance evaluator based on a classifier to obtain an evaluation result, and the evaluation result is used to represent a noise reduction performance level label.
[0028] For example, according to the building curtain wall sound insulation performance detection system of the embodiment of the present disclosure, it also includes a training module for training the gamma correction response association module, the adaptive attention weight fusion network and the classifier-based noise reduction performance evaluator.
[0029] For example, according to the building curtain wall sound insulation performance detection system of the embodiment of the present disclosure, the training module includes:
[0030] A training data acquisition unit, used to acquire training data, wherein the training data includes a time series of training noise intensity of a noise source acquired by a first acoustic device, and a time series of training sound intensity after sound insulation of a building curtain wall acquired by a second acoustic device;
[0031] A training noise intensity data time series arrangement unit, used to arrange the time series of the training noise intensity and the time series of the training sound intensity into a training noise intensity time series input vector and a training sound intensity time series input vector respectively according to the time dimension;
[0032] A training noise time series vector segmentation unit, used for performing vector segmentation on the training noise intensity time series input vector and the training sound intensity time series input vector at the same time scale to obtain a sequence of training noise intensity local time series input vectors and a sequence of training sound intensity local time series input vectors;
[0033] A training noise reduction effect local temporal response feature extraction unit, used for inputting each group of corresponding training noise intensity local temporal input vectors and training sound intensity local temporal input vectors in the sequence of training noise intensity local temporal input vectors and the sequence of training sound intensity local temporal input vectors into the gamma correction response association module to obtain a sequence of training noise reduction effect local temporal response feature vectors;
[0034] A training noise reduction effect full temporal fusion feature representation unit, used for passing the sequence of the training noise reduction effect local temporal response feature vectors through the adaptive attention weight fusion network to obtain a training noise reduction effect full temporal fusion feature vector;
[0035] A loss calculation unit, used for passing the training denoising effect full time series fusion feature vector through the classifier-based denoising performance evaluator to obtain a classification loss function value; and
[0036] A loss training unit is used to train the gamma correction response association module, the adaptive attention weight fusion network and the classifier-based noise reduction performance evaluator based on the classification loss function value, wherein in each round of iteration of the training, the full temporal fusion feature vector of the training noise reduction effect is corrected.
[0037] The embodiment of the present disclosure also provides a method for detecting the sound insulation performance of a building curtain wall, which comprises:
[0038] Acquire a time series of noise intensity of a noise source collected by a first acoustic device;
[0039] Obtaining a time series of sound intensity collected by a second acoustic device after sound insulation of the building curtain wall;
[0040] Arranging the time series of the noise intensity and the time series of the sound intensity into a noise intensity time series input vector and a sound intensity time series input vector respectively according to the time dimension;
[0041] Performing vector segmentation on the noise intensity time series input vector and the sound intensity time series input vector at the same time scale to obtain a sequence of noise intensity local time series input vectors and a sequence of sound intensity local time series input vectors;
[0042] Inputting each corresponding group of the noise intensity local time series input vector and the sound intensity local time series input vector in the sequence of the noise intensity local time series input vector and the sequence of the sound intensity local time series input vector into the gamma correction response association module to obtain a sequence of noise reduction effect local time series response feature vectors;
[0043] Performing adaptive attention weight allocation fusion processing on the sequence of local temporal response feature vectors of the noise reduction effect to obtain a full temporal fusion feature of the noise reduction effect; and
[0044] Based on the full temporal fusion characteristics of the noise reduction effect, a noise reduction performance level label is determined.
[0045] For example, according to the method for detecting the sound insulation performance of a building curtain wall in an embodiment of the present disclosure, each group of corresponding noise intensity local time series input vectors and sound intensity local time series input vectors in the sequence of the noise intensity local time series input vectors and the sequence of the sound intensity local time series input vectors are respectively input into a gamma correction response association module to obtain a sequence of local time series response feature vectors of the noise reduction effect, including:
[0046] Each group of corresponding noise intensity local time series input vectors and sound intensity local time series input vectors in the sequence of the noise intensity local time series input vectors and the sequence of the sound intensity local time series input vectors is input into the gamma correction response association module for processing using the following gamma correction response association formula to obtain the sequence of the noise reduction effect local time series response feature vectors; wherein the gamma correction response association formula is:
[0047]
[0048] in, and The first sequence of each corresponding noise intensity local time series input vector and sound intensity local time series input vector in the sequence of the noise intensity local time series input vector and the sequence of the sound intensity local time series input vector are respectively The feature values at each position, A, B, C and D are the hyperparameters for adjustment. is the first of the local time series response feature vectors of the noise reduction effect in the sequence of the local time series response feature vectors of the noise reduction effect. The feature value at each position.
[0049] For example, according to the building curtain wall sound insulation performance detection method of the embodiment of the present disclosure, the sequence of the local time series response feature vector of the noise reduction effect is subjected to adaptive attention weight allocation fusion processing to obtain the full time series fusion feature of the noise reduction effect, including:
[0050] The sequence of the local temporal response feature vectors of the noise reduction effect is passed through an adaptive attention weight fusion network to obtain a full temporal fusion feature vector of the noise reduction effect as the full temporal fusion feature of the noise reduction effect.
[0051] According to the building curtain wall sound insulation performance detection system and method of the embodiment of the present disclosure, the noise intensity of the noise source is monitored and collected in real time through the first acoustic device, and the sound intensity after the building curtain wall is soundproofed is monitored and collected in real time through the second acoustic device, and an artificial intelligence-based data processing and analysis algorithm is introduced at the back end to perform a time series coordination and responsiveness correlation analysis between the noise intensity of the noise source and the sound intensity after the building curtain wall is soundproofed, so as to perform sound insulation performance detection of the building curtain wall, thereby judging the noise reduction performance level of the building curtain wall. In this way, intelligent automatic detection of the sound insulation performance of the building curtain wall can be realized, which not only avoids the artificial errors and inefficiency problems introduced by the traditional method, but also can test the noise reduction performance of the building curtain wall under different noise time series conditions, so as to improve the efficiency and accuracy of the detection of the sound insulation performance of the building curtain wall. BRIEF DESCRIPTION OF THE DRAWINGS
[0052] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the following briefly introduces the drawings of the embodiments of the present disclosure. Obviously, the drawings described below only relate to some embodiments of the present disclosure, and are not intended to limit the present disclosure.
[0053] Figure 1 A schematic diagram of the application architecture of a building curtain wall sound insulation performance detection system in an embodiment of the present disclosure is shown;
[0054] Figure 2 A schematic diagram of the structure of a building curtain wall sound insulation performance detection system in an embodiment of the present disclosure is shown;
[0055] Figure 3 A flow chart of a method for detecting sound insulation performance of a building curtain wall in an embodiment of the present disclosure is shown;
[0056] Figure 4 An application scenario diagram of the building curtain wall sound insulation performance detection system in an embodiment of the present disclosure is shown; and
[0057] Figure 5A schematic diagram of a storage medium according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION
[0058] The technical solutions in the embodiments of the present disclosure will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present disclosure, rather than all embodiments. Based on the embodiments of the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work also fall within the scope of protection of the present disclosure.
[0059] The terms used in this specification are those general terms currently widely used in the art in consideration of the functions of the present disclosure, but these terms may vary according to the intention of a person of ordinary skill in the art, precedents, or new technologies in the art. In addition, specific terms may be selected by the applicant, and in this case, their detailed meanings will be described in the detailed description of the present disclosure. Therefore, the terms used in the specification should not be understood as simple names, but rather as an overall description based on the meaning of the terms and the present disclosure.
[0060] Although the present disclosure makes various references to certain modules in the system according to embodiments of the present disclosure, any number of different modules can be used and run on a user terminal and / or server. The modules are illustrative only, and different aspects of the systems and methods can use different modules.
[0061] Flowcharts are used in this disclosure to illustrate the operations performed by the system according to the embodiments of the present disclosure. It should be understood that the preceding or following operations are not necessarily performed precisely in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. At the same time, other operations may also be added to these processes, or one or more operations may be removed from these processes.
[0062] Figure 1 A schematic diagram of the application architecture of a building curtain wall sound insulation performance detection system in an embodiment of the present disclosure is shown, including a server 100 and a terminal device 200.
[0063] The terminal device 200 and the server 100 can be connected via the Internet to achieve mutual communication. Optionally, the above-mentioned Internet uses standard communication technology and / or protocol. The Internet is usually the Internet, but it can also be any network, including but not limited to any combination of a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), a mobile, wired or wireless network, a private network or a virtual private network. In some embodiments, the data exchanged through the network is represented by technologies and / or formats including Hyper Text Markup Language (HTML), Extensible Markup Language (XML), etc. In addition, conventional encryption technologies such as Secure Socket Layer (SSL), Transport Layer Security (TLS), Virtual Private Network (VPN), Internet Protocol Security (IPsec), etc. can also be used to encrypt all or some links. In other embodiments, customized and / or dedicated data communication technologies can also be used to replace or supplement the above-mentioned data communication technologies.
[0064] The server 100 can provide various network services for the terminal device 200, wherein the server 100 can be a single server, a server cluster consisting of several servers, or a cloud computing center. Specifically, the server 100 may include a processor 110 (Center Processing Unit, CPU), a memory 120, an input device 130, and an output device 140, etc. The input device 130 may include a keyboard, a mouse, a touch screen, etc., and the output device 140 may include a display device, such as a liquid crystal display (Liquid Crystal Display, LCD), a cathode ray tube (Cathode Ray Tube, CRT), etc.
[0065] The memory 120 may include a read-only memory (ROM) and a random access memory (RAM), and provides the processor 110 with program instructions and data stored in the memory 120. In the embodiment of the present disclosure, the memory 120 may be used to store the corresponding method of the building curtain wall sound insulation performance detection system in the embodiment of the present disclosure.
[0066] The processor 110 calls the program instructions stored in the memory 120, and the processor 110 is used to execute the steps of any one of the building curtain wall sound insulation performance detection methods in the embodiments of the present disclosure according to the obtained program instructions.
[0067] In addition, the application architecture diagram in the embodiment of the present disclosure is intended to more clearly illustrate the technical solution in the embodiment of the present disclosure, and does not constitute a limitation on the technical solution provided in the embodiment of the present disclosure. Of course, for other application architectures and business applications, the technical solution provided in the embodiment of the present disclosure is also applicable to similar problems.
[0068] The following is a non-restrictive description of the building curtain wall sound insulation performance detection system provided according to at least one embodiment of the present disclosure through several examples or embodiments. As described below, different features in these specific examples or embodiments can be combined with each other without conflicting with each other to obtain new examples or embodiments, and these new examples or embodiments also fall within the scope of protection of the present disclosure.
[0069] In response to the above technical problems, the technical concept of the present application is to monitor and collect the noise intensity of the noise source in real time through the first acoustic device, monitor and collect the sound intensity after the building curtain wall is soundproofed in real time through the second acoustic device, and introduce artificial intelligence-based data processing and analysis algorithms at the back end to perform time series coordination and responsiveness correlation analysis between the noise intensity of the noise source and the sound intensity after the building curtain wall is soundproofed, so as to detect the sound insulation performance of the building curtain wall and judge the noise reduction performance level of the building curtain wall. In this way, intelligent automatic detection of the sound insulation performance of the building curtain wall can be realized, which not only avoids the artificial errors and inefficiency problems introduced by the traditional method, but also can test the noise reduction performance of the building curtain wall under different noise time series conditions, so as to improve the efficiency and accuracy of the detection of the sound insulation performance of the building curtain wall.
[0070] Figure 2The structural schematic diagram of the building curtain wall sound insulation performance detection system 800 in the embodiment of the present disclosure is shown. The building curtain wall sound insulation performance detection system 800 includes: a noise source noise intensity data acquisition module 810, which is used to obtain the time series of the noise intensity of the noise source collected by the first acoustic device; a noise intensity data acquisition module 820 after sound insulation, which is used to obtain the time series of the sound intensity collected by the second acoustic device after the building curtain wall is soundproofed; a noise intensity data time series arrangement module 830, which is used to arrange the time series of the noise intensity and the time series of the sound intensity into a noise intensity time series input vector and a sound intensity time series input vector according to the time dimension respectively; a noise time series vector segmentation module 840, which is used to perform vector segmentation on the noise intensity time series input vector and the sound intensity time series input vector at the same time scale to obtain a noise intensity local time series input vector A noise reduction effect local temporal response feature extraction module 850 is used to input each group of corresponding noise intensity local temporal input vectors and sound intensity local temporal input vectors in the sequence of noise intensity local temporal input vectors and the sequence of sound intensity local temporal input vectors into a gamma correction response association module to obtain a sequence of local temporal response feature vectors of noise reduction effect; a noise reduction effect full temporal fusion feature representation module 860 is used to perform adaptive attention weight allocation fusion processing on the sequence of local temporal response feature vectors of noise reduction effect to obtain the full temporal fusion feature of noise reduction effect; and a noise reduction performance level detection module 870 is used to determine the noise reduction performance level label based on the full temporal fusion feature of noise reduction effect.
[0071] Specifically, in the technical solution of the present application, first, the time series of the noise intensity of the noise source collected by the first acoustic device is obtained, and the time series of the sound intensity after the building curtain wall sound insulation is obtained by the second acoustic device. Next, considering that the noise intensity of the noise source and the sound intensity after the building curtain wall sound insulation will continue to change with the passage of time, in order to ensure that the noise intensity and the noise intensity data at the same time point are aligned, so as to facilitate the subsequent analysis of the noise reduction performance of the building curtain wall, in the technical solution of the present application, it is necessary to arrange the time series of the noise intensity and the time series of the sound intensity according to the time dimension as a noise intensity time series input vector and a sound intensity time series input vector respectively. In this way, it is conducive to discovering the time series correlation and change regularity between the noise intensity of the noise source and the sound intensity after the building curtain wall sound insulation.
[0072] Then, considering that when noise monitoring is performed to judge the noise reduction performance of the building curtain wall, the amount of time series data of the noise intensity of the noise source and the sound intensity after the building curtain wall is soundproofed may be too large, and it is difficult to capture the subtle changes in the sound and the time series related pattern feature information. Therefore, in order to be able to perform more detailed and comprehensive time series analysis and associated coding of noise intensity and sound intensity, so as to more accurately analyze the sound insulation performance of the building curtain wall, in the technical solution of the present application, the noise intensity time series input vector and the sound intensity time series input vector are further vector segmented at the same time scale to obtain a sequence of local time series input vectors of noise intensity and a sequence of local time series input vectors of sound intensity. By segmenting the time series input vectors of noise intensity and sound intensity into a sequence of local time series input vectors according to the same time scale, the noise and sound data in the same time period can be made to correspond, which is convenient for local comparison and analysis. This helps to discover the correlation and regularity between noise and sound data, which is conducive to analyzing the sound insulation performance of the building curtain wall in different time periods and under different noise time series conditions.
[0073] Then, considering that there is a responsive correlation performance based on the sound insulation effect of the building curtain wall between the noise intensity of the noise source and the sound intensity after the building curtain wall is soundproofed, this responsive correlation feature is nonlinear and difficult to effectively characterize and capture. Based on this, in the technical solution of the present application, each group of corresponding noise intensity local time series input vectors and sound intensity local time series input vectors in the sequence of the noise intensity local time series input vectors and the sequence of the sound intensity local time series input vectors are further input into the gamma correction response association module to obtain a sequence of local time series response feature vectors of the noise reduction effect. Through the processing of the gamma correction response association module, the nonlinear responsive correlation pattern and characteristics between the noise intensity of the noise source and the sound intensity after the building curtain wall is soundproofed in each local time period can be captured, so as to perform local time series characterization of the noise reduction effect in different local time periods, so as to more accurately evaluate the sound insulation performance of the building curtain wall.
[0074] Correspondingly, the noise reduction effect local time series response feature extraction module 850 is used to: input each group of corresponding noise intensity local time series input vectors and sound intensity local time series input vectors in the sequence of the noise intensity local time series input vectors and the sequence of the sound intensity local time series input vectors into the gamma correction response association module for processing using the following gamma correction response association formula to obtain the sequence of the noise reduction effect local time series response feature vectors; wherein the gamma correction response association formula is:
[0075]
[0076] in, and The first and second sequences of the noise intensity local time series input vectors and the sound intensity local time series input vectors in each group are respectively the sequence of the noise intensity local time series input vectors and the sequence of the sound intensity local time series input vectors. The feature values at each position, A, B, C and D are the hyperparameters for adjustment. is the first of the local time series response feature vectors of the noise reduction effect in the sequence of the local time series response feature vectors of the noise reduction effect. The feature value at each position.
[0077] Further, considering that in the actual noise monitoring and noise reduction performance detection process of building curtain walls, the gamma-corrected responsiveness correlation characteristics between the noise intensity in different local time periods and the sound intensity after sound insulation have a correlation relationship based on the whole time domain. That is to say, the temporal characterization characteristics of the noise reduction effect in different time periods or local areas often contain the noise reduction effect information under different noise conditions, and in the actual process of judging the noise reduction performance of building curtain walls, the local temporal characteristics of the noise reduction effect under these different noise conditions have different importance. Based on this, in the technical solution of the present application, the sequence of the local temporal response feature vector of the noise reduction effect is further passed through an adaptive attention weight fusion network to obtain the full temporal fusion feature vector of the noise reduction effect. Through the processing of the adaptive attention weight fusion network, the importance of the temporal characteristics of the noise reduction effect in different time periods or local areas can be automatically learned according to the sequence of the input local temporal response feature vector of the noise reduction effect, so as to adaptively allocate weights according to the importance, so as to more effectively integrate the noise reduction effect features in different local time periods. In other words, through the adaptive attention mechanism, the network can pay more attention to the part that contributes to the noise reduction effect in the whole time domain, and improve the effect of feature fusion and noise reduction performance. In this way, it helps to extract a more comprehensive and accurate full-time domain feature representation of the noise reduction effect of the building curtain wall, providing stronger support for further analysis and evaluation.
[0078] Correspondingly, the noise reduction effect full temporal fusion feature characterization module 860 is used to: pass the sequence of the noise reduction effect local temporal response feature vectors through an adaptive attention weight fusion network to obtain the noise reduction effect full temporal fusion feature vector as the noise reduction effect full temporal fusion feature.
[0079] Specifically, the noise reduction effect full temporal fusion feature characterization module 860 is used to: process the sequence of the local temporal response feature vectors of the noise reduction effect through the adaptive attention weight fusion network with the following adaptive attention fusion formula to obtain the noise reduction effect full temporal fusion feature vector; wherein the adaptive attention fusion formula is:
[0080]
[0081]
[0082]
[0083] in, is the first in the sequence of local temporal response feature vectors of the noise reduction effect The local time series response feature vector of the noise reduction effect, and Represent the weight coefficient vector and weight coefficient matrix respectively, is the offset vector, express function, For the The attention score of the local temporal response feature vector of the noise reduction effect, and are all hyperparameters, express function, is the number of vectors in the sequence of the local temporal response feature vectors of the noise reduction effect, It is the full temporal fusion feature vector of the noise reduction effect.
[0084] Furthermore, the full-time fusion feature vector of the noise reduction effect is passed through a noise reduction performance evaluator based on a classifier to obtain an evaluation result, and the evaluation result is used to represent the noise reduction performance level label. In other words, the full-time weighted fusion characterization feature of the noise reduction effect of the building curtain wall is used for classification processing to detect the sound insulation performance of the building curtain wall, thereby determining the noise reduction performance level of the building curtain wall. In this way, intelligent automatic detection of the sound insulation performance of the building curtain wall can be achieved, which not only avoids the artificial errors and inefficiency problems introduced by the traditional method, but also can test the noise reduction performance of the building curtain wall under different noise time series conditions, so as to improve the efficiency and accuracy of the detection of the sound insulation performance of the building curtain wall.
[0085] Correspondingly, the noise reduction performance level detection module 870 is used to: pass the noise reduction effect full time series fusion feature vector through a classifier-based noise reduction performance evaluator to obtain an evaluation result, and the evaluation result is used to represent the noise reduction performance level label.
[0086] It should be understood that the role of the classifier is to use the given categories and known training data to learn classification rules and classifiers, and then classify (or predict) unknown data. Logistic regression, SVM, etc. are often used to solve binary classification problems. For multi-class classification problems, logistic regression or SVM can also be used, but multiple binary classifications are required to form a multi-classification, but this is prone to errors and inefficient. Commonly used multi-classification methods include the Softmax classification function.
[0087] Furthermore, in the technical solution of the present application, the building curtain wall sound insulation performance detection system also includes a training module for training the gamma correction response association module, the adaptive attention weight fusion network and the classifier-based noise reduction performance evaluator.
[0088] In one example, the training module includes: a training data acquisition unit, which is used to acquire training data, wherein the training data includes a time series of training noise intensities of a noise source collected by a first acoustic device, and a time series of training sound intensities collected by a second acoustic device after sound insulation of a building curtain wall; a training noise intensity data time series arrangement unit, which is used to arrange the time series of training noise intensities and the time series of training sound intensities according to the time dimension into a training noise intensity time series input vector and a training sound intensity time series input vector, respectively; a training noise time series vector segmentation unit, which is used to perform vector segmentation on the training noise intensity time series input vector and the training sound intensity time series input vector at the same time scale to obtain a sequence of training noise intensity local time series input vectors and a sequence of training sound intensity local time series input vectors; a training noise reduction effect local time series response feature extraction unit, which is used to arrange the sequence of training noise intensity local time series input vectors and the sequence of training sound intensity local time series input vectors into a training noise reduction effect local time series response feature extraction unit. Each group of corresponding training noise intensity local time series input vectors and training sound intensity local time series input vectors in the sequence of intensity local time series input vectors is respectively input into the gamma correction response association module to obtain a sequence of training noise reduction effect local time series response feature vectors; a training noise reduction effect full time series fusion feature representation unit is used to pass the sequence of training noise reduction effect local time series response feature vectors through the adaptive attention weight fusion network to obtain a training noise reduction effect full time series fusion feature vector; a loss calculation unit is used to pass the training noise reduction effect full time series fusion feature vector through the classifier-based noise reduction performance evaluator to obtain a classification loss function value; and a loss training unit is used to train the gamma correction response association module, the adaptive attention weight fusion network and the classifier-based noise reduction performance evaluator based on the classification loss function value, wherein, in each round of iteration of the training, the training noise reduction effect full time series fusion feature vector is corrected.
[0089] In the technical solution described above, the sequence of local temporal response feature vectors of the training noise reduction effect respectively expresses the temporal correlation characteristics of the training noise intensity and the training sound intensity in the local time domain based on the gamma correction response. Therefore, after the sequence of local temporal response feature vectors of the training noise reduction effect passes through the adaptive attention weight fusion network, the obtained full temporal fusion feature vector of the training noise reduction effect will be subjected to local time domain adaptive attention weighted fusion based on the overall feature distribution of the local time domain under the global time domain, so that the full temporal fusion feature vector of the training noise reduction effect has the temporal correlation feature expression complexity of the differentiated local time domain distribution related to the local-global time domain, that is, the full temporal fusion feature vector of the training noise reduction effect will have rich geometric shape details in the high-dimensional temporal feature space, so that there is category representation distortion when it is classified by the classifier to map to the class probability domain, that is, it affects the category representation of the full temporal fusion feature vector of the training noise reduction effect for the category predetermined by the classifier, thereby reducing the accuracy of the classification result.
[0090] Based on this, the applicant of the present application corrects the full-time series fusion feature vector of the training denoising effect each time the full-time series fusion feature vector of the training denoising effect is classified and iterated through the classifier.
[0091] Accordingly, in one example, in each iteration of the training, the training denoising effect full time series fusion feature vector is corrected by the following correction formula to obtain a corrected training denoising effect full time series fusion feature vector; wherein the correction formula is
[0092]
[0093] in, is the full time series fusion feature vector of the training denoising effect, is the full time series fusion feature vector of the training denoising effect The class probability value obtained by the classifier, is the full time series fusion feature vector of the noise reduction effect No. eigenvalues, is the full time series fusion feature vector of the training denoising effect The mean of all eigenvalues of Represents the full time series fusion feature vector of the training denoising effect The one-norm of is the weight hyperparameter, represents an exponential operation of a value, wherein the exponential operation of the value represents calculating the value of a natural exponential function with the value as a power, is the first time series fusion feature vector of the denoising effect after correction training feature values.
[0094] Specifically, in the training denoising effect full time series fusion feature vector While maintaining the geometric details of the high-dimensional feature manifold under the class probability mapping, the eigenvalue details are fused with the feature vector in full time series relative to the training denoising effect by means of class probability distortion consultation. The variational representation of the feature set is used to edit the shape attributes, thereby integrating the training denoising effect into the full time series fusion feature vector The class probability distortion is mapped to the latent class space feature representation, and then the basic low-rank constrained latent inversion representation is supplemented to perform consultation fusion to bridge the gap between the edited consultation representation and the original geometric shape details, thereby improving the training denoising effect. Full temporal fusion feature vector The high-dimensional features of the class representation can be improved to improve the accuracy of the classification results. In this way, the sound insulation performance of the building curtain wall can be tested more accurately, thereby automatically judging the noise reduction performance level of the building curtain wall.
[0095] Based on the above embodiments, see Figure 3 FIG. 1 is a flow chart of a method for detecting the sound insulation performance of a building curtain wall in an embodiment of the present disclosure. For example, the method for detecting the sound insulation performance of a building curtain wall can be executed by a server, which can be Figure 1 The server 100 shown in FIG. Figure 3 As shown, according to the embodiment of the present disclosure, the method for detecting the sound insulation performance of the building curtain wall includes the following steps: S510, obtaining the time series of the noise intensity of the noise source collected by the first acoustic device; S520, obtaining the time series of the sound intensity after the building curtain wall sound insulation collected by the second acoustic device; S530, arranging the time series of the noise intensity and the time series of the sound intensity into a noise intensity time series input vector and a sound intensity time series input vector according to the time dimension respectively; S540, vector segmenting the noise intensity time series input vector and the sound intensity time series input vector at the same time scale to obtain a noise intensity local time series input vector The sequence of local temporal input vectors of sound intensity and the sequence of local temporal input vectors of sound intensity; S550, respectively input each group of corresponding local temporal input vectors of noise intensity and local temporal input vectors of sound intensity in the sequence of local temporal input vectors of noise intensity and the sequence of local temporal input vectors of sound intensity into the gamma correction response association module to obtain a sequence of local temporal response feature vectors of noise reduction effect; S560, perform adaptive attention weight allocation fusion processing on the sequence of local temporal response feature vectors of noise reduction effect to obtain the full temporal fusion feature of noise reduction effect; and, S570, determine the noise reduction performance level label based on the full temporal fusion feature of noise reduction effect.
[0096] In one example, in the above-mentioned building curtain wall sound insulation performance detection method, each group of corresponding noise intensity local time series input vectors and sound intensity local time series input vectors in the sequence of the noise intensity local time series input vectors and the sequence of the sound intensity local time series input vectors are respectively input into a gamma correction response association module to obtain a sequence of local time series response feature vectors of noise reduction effect, including: inputting each group of corresponding noise intensity local time series input vectors and sound intensity local time series input vectors in the sequence of the noise intensity local time series input vectors and the sequence of the sound intensity local time series input vectors into the gamma correction response association module and processing them with the following gamma correction response association formula to obtain the sequence of local time series response feature vectors of noise reduction effect; wherein, the gamma correction response association formula is:
[0097]
[0098] in, and The first sequence of each corresponding noise intensity local time series input vector and sound intensity local time series input vector in the sequence of the noise intensity local time series input vector and the sequence of the sound intensity local time series input vector are respectively The feature values at each position, A, B, C and D are the hyperparameters for adjustment. is the first of the local time series response feature vectors of the noise reduction effect in the sequence of the local time series response feature vectors of the noise reduction effect. The feature value at each position.
[0099] In one example, in the above-mentioned building curtain wall sound insulation performance detection method, the sequence of the local temporal response feature vectors of the noise reduction effect is subjected to adaptive attention weight allocation fusion processing to obtain the full temporal fusion feature of the noise reduction effect, including: passing the sequence of the local temporal response feature vectors of the noise reduction effect through an adaptive attention weight fusion network to obtain the full temporal fusion feature vector of the noise reduction effect as the full temporal fusion feature of the noise reduction effect.
[0100] Here, those skilled in the art will appreciate that the specific operations of each step in the above-mentioned building curtain wall sound insulation performance detection method have been referenced above. Figure 2 The description of the building curtain wall sound insulation performance detection system 800 has been introduced in detail, and therefore, its repeated description will be omitted.
[0101] Figure 4 FIG. 1 is an application scenario diagram of the building curtain wall sound insulation performance detection system according to an embodiment of the present disclosure. Figure 4 As shown, in this application scenario, first, a time series of noise intensity of a noise source collected by a first acoustic device is obtained (for example, Figure 4D1 shown in FIG), and the time series of the sound intensity after the building curtain wall sound insulation collected by the second acoustic device (for example, Figure 4 Then, the time series of the noise intensity and the time series of the sound intensity are input to a server (for example, Figure 4 In S) as shown in , the server is capable of using the building curtain wall sound insulation performance detection algorithm to process the time series of the noise intensity and the time series of the sound intensity to obtain an evaluation result for representing a noise reduction performance level label.
[0102] Based on the above embodiments, another exemplary embodiment of an electronic device is also provided in the embodiments of the present disclosure. In some possible implementations, the electronic device in the embodiments of the present disclosure may include a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor can implement the steps of the building curtain wall sound insulation performance detection method in the above embodiments when executing the program.
[0103] For example, the electronic device is disclosed in the present invention Figure 1 Taking the server 100 in the example as an example, the processor in the electronic device is the processor 110 in the server 100, and the memory in the electronic device is the memory 120 in the server 100.
[0104] An embodiment of the present disclosure also provides a computer-readable storage medium. Figure 5 FIG. 1 is a schematic diagram of a computer-readable storage medium 1000 according to an embodiment of the present disclosure. Figure 5 As shown, the computer-readable storage medium 1000 stores computer-executable instructions 1001. When the computer-executable instructions 1001 are executed by the processor, the building curtain wall sound insulation performance detection method according to the embodiment of the present disclosure described with reference to the above figures can be executed. The computer-readable storage medium includes, but is not limited to, for example, volatile memory and / or non-volatile memory. The volatile memory may include, for example, a random access memory (RAM) and / or a cache memory (cache), etc. The non-volatile memory may include, for example, a read-only memory (ROM), a hard disk, a flash memory, etc.
[0105] The embodiments of the present disclosure also provide a computer program product or a computer program, which includes a computer instruction stored in a computer-readable storage medium. The processor of the computer device reads the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes the building curtain wall sound insulation performance detection method according to the embodiments of the present disclosure.
[0106] Those skilled in the art will appreciate that the contents disclosed in this disclosure may be subject to various variations and improvements. For example, the various devices or components described above may be implemented by hardware, or by software, firmware, or a combination of some or all of the three.
[0107] In addition, although the present disclosure makes various references to certain units in the system according to embodiments of the present disclosure, any number of different units may be used and run on the client and / or server. The units are illustrative only, and different aspects of the systems and methods may use different units.
[0108] Those skilled in the art will appreciate that all or part of the steps in the above method can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk. Optionally, all or part of the steps in the above embodiment can also be implemented using one or more integrated circuits. Accordingly, each module / unit in the above embodiment can be implemented in the form of hardware or in the form of a software functional module. The present disclosure is not limited to any particular form of combination of hardware and software.
[0109] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present disclosure belongs. It should also be understood that terms such as those defined in common dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology and should not be interpreted in an idealized or highly formal sense, unless explicitly defined as such herein.
[0110] The above is an explanation of the present disclosure and should not be considered as a limitation thereof. Although several exemplary embodiments of the present disclosure have been described, it will be readily understood by those skilled in the art that many modifications may be made to the exemplary embodiments without departing from the novel teachings and advantages of the present disclosure. Therefore, all such modifications are intended to be included within the scope of the present disclosure as defined in the claims. It should be understood that the above is an explanation of the present disclosure and should not be considered as being limited to the specific embodiments disclosed, and modifications to the disclosed embodiments and other embodiments are intended to be included within the scope of the appended claims. The present disclosure is defined by the claims and their equivalents.
Claims
1. A building curtain wall sound insulation performance detection system, It is characterized in that include: A noise source noise intensity data acquisition module, used to acquire a time series of noise intensity of the noise source collected by the first acoustic device; A noise intensity data acquisition module after sound insulation, used to obtain the time series of the sound intensity after sound insulation of the building curtain wall collected by the second acoustic device; A noise intensity data time series arrangement module, used for arranging the time series of the noise intensity and the time series of the sound intensity into a noise intensity time series input vector and a sound intensity time series input vector respectively according to the time dimension; A noise time series vector segmentation module, used for performing vector segmentation on the noise intensity time series input vector and the sound intensity time series input vector at the same time scale to obtain a sequence of noise intensity local time series input vectors and a sequence of sound intensity local time series input vectors; A noise reduction effect local time series response feature extraction module, used for inputting each group of corresponding noise intensity local time series input vectors and sound intensity local time series input vectors in the sequence of the noise intensity local time series input vectors and the sequence of the sound intensity local time series input vectors into a gamma correction response association module to obtain a sequence of noise reduction effect local time series response feature vectors; A noise reduction effect full-time fusion feature characterization module is used to perform adaptive attention weight allocation fusion processing on the sequence of the local time series response feature vectors of the noise reduction effect to obtain the full-time fusion feature of the noise reduction effect; as well as A noise reduction performance level detection module, used to determine a noise reduction performance level label based on the full temporal fusion characteristics of the noise reduction effect; Wherein, the noise reduction effect local time series response feature extraction module is used to: Each group of corresponding noise intensity local time series input vectors and sound intensity local time series input vectors in the sequence of the noise intensity local time series input vectors and the sequence of the sound intensity local time series input vectors is input into the gamma correction response association module for processing using the following gamma correction response association formula to obtain the sequence of the noise reduction effect local time series response feature vectors; wherein the gamma correction response association formula is: in, and The first sequence of each corresponding noise intensity local time series input vector and sound intensity local time series input vector in the sequence of the noise intensity local time series input vector and the sequence of the sound intensity local time series input vector are respectively The feature values at each position, A, B, C and D are the hyperparameters for adjustment. is the first of the local time series response feature vectors of the noise reduction effect in the sequence of the local time series response feature vectors of the noise reduction effect. The eigenvalues of the positions; The noise reduction effect full-time fusion feature characterization module is used to: The sequence of the local temporal response feature vectors of the noise reduction effect is passed through an adaptive attention weight fusion network to obtain a full temporal fusion feature vector of the noise reduction effect as the full temporal fusion feature of the noise reduction effect; The noise reduction effect full-time fusion feature characterization module is used to: The sequence of the local temporal response feature vectors of the noise reduction effect is processed by the adaptive attention weight fusion network using the following adaptive attention fusion formula to obtain the full temporal fusion feature vector of the noise reduction effect; wherein the adaptive attention fusion formula is: in, is the first in the sequence of local temporal response feature vectors of the noise reduction effect The local time series response feature vector of the noise reduction effect, and Represent the weight coefficient vector and weight coefficient matrix respectively, is the offset vector, express function, For the The attention score of the local temporal response feature vector of the noise reduction effect, and are all hyperparameters, express function, is the number of vectors in the sequence of the local temporal response feature vectors of the noise reduction effect, It is the full temporal fusion feature vector of the noise reduction effect.
2. The building curtain wall sound insulation performance detection system according to claim 1, It is characterized in that The noise reduction performance level detection module is used to: The full-time fusion feature vector of the noise reduction effect is passed through a noise reduction performance evaluator based on a classifier to obtain an evaluation result, and the evaluation result is used to represent a noise reduction performance level label.
3. The building curtain wall sound insulation performance detection system according to claim 2, It is characterized in that It also includes a training module for training the gamma correction response association module, the adaptive attention weight fusion network and the classifier-based noise reduction performance evaluator.
4. The building curtain wall sound insulation performance detection system according to claim 3, It is characterized in that The training module comprises: A training data acquisition unit, used to acquire training data, wherein the training data includes a time series of training noise intensity of a noise source acquired by a first acoustic device, and a time series of training sound intensity after sound insulation of a building curtain wall acquired by a second acoustic device; A training noise intensity data time series arrangement unit, used to arrange the time series of the training noise intensity and the time series of the training sound intensity into a training noise intensity time series input vector and a training sound intensity time series input vector respectively according to the time dimension; A training noise time series vector segmentation unit, used for performing vector segmentation on the training noise intensity time series input vector and the training sound intensity time series input vector at the same time scale to obtain a sequence of training noise intensity local time series input vectors and a sequence of training sound intensity local time series input vectors; A training noise reduction effect local temporal response feature extraction unit, used for inputting each group of corresponding training noise intensity local temporal input vectors and training sound intensity local temporal input vectors in the sequence of training noise intensity local temporal input vectors and the sequence of training sound intensity local temporal input vectors into the gamma correction response association module to obtain a sequence of training noise reduction effect local temporal response feature vectors; A training noise reduction effect full temporal fusion feature representation unit, used for passing the sequence of the training noise reduction effect local temporal response feature vectors through the adaptive attention weight fusion network to obtain a training noise reduction effect full temporal fusion feature vector; A loss calculation unit, used for passing the training denoising effect full time series fusion feature vector through the classifier-based denoising performance evaluator to obtain a classification loss function value; and A loss training unit is used to train the gamma correction response association module, the adaptive attention weight fusion network and the classifier-based noise reduction performance evaluator based on the classification loss function value, wherein in each round of iteration of the training, the full temporal fusion feature vector of the training noise reduction effect is corrected.
5. A method for testing the sound insulation performance of building curtain walls. It is characterized in that include: Acquire a time series of noise intensity of a noise source collected by a first acoustic device; Obtaining a time series of sound intensity collected by a second acoustic device after sound insulation of the building curtain wall; Arranging the time series of the noise intensity and the time series of the sound intensity into a noise intensity time series input vector and a sound intensity time series input vector respectively according to the time dimension; Performing vector segmentation on the noise intensity time series input vector and the sound intensity time series input vector at the same time scale to obtain a sequence of noise intensity local time series input vectors and a sequence of sound intensity local time series input vectors; Inputting each corresponding group of the noise intensity local time series input vector and the sound intensity local time series input vector in the sequence of the noise intensity local time series input vector and the sequence of the sound intensity local time series input vector into the gamma correction response association module to obtain a sequence of noise reduction effect local time series response feature vectors; Performing adaptive attention weight allocation fusion processing on the sequence of local temporal response feature vectors of the noise reduction effect to obtain a full temporal fusion feature of the noise reduction effect; as well as Determining a noise reduction performance level label based on the full temporal fusion characteristics of the noise reduction effect; The steps of inputting each corresponding group of the noise intensity local time series input vector and the sound intensity local time series input vector in the sequence of the noise intensity local time series input vector and the sequence of the sound intensity local time series input vector into the gamma correction response association module to obtain a sequence of the noise reduction effect local time series response feature vectors include: Each group of corresponding noise intensity local time series input vectors and sound intensity local time series input vectors in the sequence of the noise intensity local time series input vectors and the sequence of the sound intensity local time series input vectors is input into the gamma correction response association module for processing using the following gamma correction response association formula to obtain the sequence of the noise reduction effect local time series response feature vectors; wherein the gamma correction response association formula is: in, and The first sequence of each corresponding noise intensity local time series input vector and sound intensity local time series input vector in the sequence of the noise intensity local time series input vector and the sequence of the sound intensity local time series input vector are respectively The feature values at each position, A, B, C and D are the hyperparameters for adjustment. is the first of the local time series response feature vectors of the noise reduction effect in the sequence of the local time series response feature vectors of the noise reduction effect. The eigenvalues of the positions; The sequence of the local temporal response feature vectors of the noise reduction effect is subjected to adaptive attention weight allocation fusion processing to obtain the full temporal fusion feature of the noise reduction effect, including: The sequence of the local temporal response feature vectors of the noise reduction effect is passed through an adaptive attention weight fusion network to obtain a full temporal fusion feature vector of the noise reduction effect as the full temporal fusion feature of the noise reduction effect; The sequence of the local temporal response feature vectors of the noise reduction effect is subjected to adaptive attention weight allocation fusion processing to obtain the full temporal fusion feature of the noise reduction effect, including: The sequence of the local temporal response feature vectors of the noise reduction effect is processed by the adaptive attention weight fusion network using the following adaptive attention fusion formula to obtain the full temporal fusion feature vector of the noise reduction effect; wherein the adaptive attention fusion formula is: in, is the first in the sequence of local temporal response feature vectors of the noise reduction effect The local time series response feature vector of the noise reduction effect, and Represent the weight coefficient vector and weight coefficient matrix respectively, is the offset vector, express function, For the The attention score of the local temporal response feature vector of the noise reduction effect, and are all hyperparameters, express function, is the number of vectors in the sequence of the local temporal response feature vectors of the noise reduction effect, It is the full temporal fusion feature vector of the noise reduction effect.
