Intelligent signal processing method and system based on augmented reality

Abstract

The present invention belongs to the field of building monitoring signal processing, and in particular, relates to an intelligent signal processing method and system based on augmented reality. The method firstly uses a built-in bilateral filtering algorithm to preprocess vibration or settlement signals and saves them through an edge server. Secondly, a signal dynamic compensation model constructed based on an RLS filter performs secondary filtering and compensation on the signals. Thirdly, a dynamic model is constructed according to three-dimensional building data, and the filtered signal is mapped to the model to simulate the vibration or settlement process. The assessment model constructed by a fuzzy algorithm is used to assess the vibration or settlement risk level and mark it on the three-dimensional model. Finally, the model is displayed through an AR device, and a processing strategy library is automatically called according to the risk level to achieve risk warning and strategy display. The present invention not only improves the accuracy and transmission efficiency of the transmission signal, but also improves the accuracy and real-time performance of risk assessment.

Description

Technical Field

[0001] The present invention belongs to the field of building monitoring signal processing, and in particular relates to an intelligent signal processing method and system based on augmented reality. Background Art

[0002] With the rapid development of IoT technology, hundreds of millions of sensor devices are deployed around the world, especially in the fields of building monitoring, smart cities, industrial automation, etc. These devices continuously generate massive amounts of real-time data. These data contain important information about the status of buildings, environmental changes and potential risks, but also bring unprecedented data processing challenges. Real-time, accuracy and efficiency have become key requirements for processing this data, especially in dynamic environments that require rapid response, such as monitoring the vibration or settlement of buildings to prevent safety risks.

[0003] For example, the patent with authorization announcement number CN111525940B discloses a wearable device, a signal processing method, a signal processing apparatus and a storage medium, wherein the wearable device includes: a glasses terminal and a processing terminal, wherein the glasses terminal includes a first electrode, and when the glasses terminal is worn, the first electrode is in contact with the human body; the processing terminal includes a second electrode, and when the processing terminal is worn, the second electrode is in contact with the human body; the glasses terminal is used to send a sensing signal through the first electrode, so that the sensing signal is transmitted through the human body to reach the second electrode; the processing terminal is used to process the sensing signal to obtain a feedback signal, and send the feedback signal through the second electrode, so that the feedback signal is transmitted through the human body to reach the first electrode.

[0004] The above existing technologies all have the following problems: the existing signal processing technology often causes large delays when processing large amounts of real-time data from IoT devices due to problems with convergence speed and processing accuracy, and cannot meet real-time requirements. In order to solve the above problems, the present invention provides an intelligent signal processing method and system based on augmented reality. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention proposes an intelligent signal processing method and system based on augmented reality. The method first uses a built-in bilateral filtering algorithm to preprocess the vibration or settlement signal and saves it through an edge server. Secondly, the signal dynamic compensation model constructed based on the RLS filter is used to perform secondary filtering and compensation on the signal. Thirdly, a dynamic model is constructed according to the three-dimensional data of the building, and the filtered signal is mapped to the model to simulate the vibration or settlement process. The evaluation model constructed by the fuzzy algorithm is used to evaluate the vibration or settlement risk level and mark it on the three-dimensional model. Finally, the model is displayed through an AR device, and the processing strategy library is automatically called according to the risk level to achieve risk warning and strategy display. The present invention not only improves the accuracy and transmission efficiency of the transmission signal, but also improves the accuracy and real-time performance of risk assessment.

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

[0007] Intelligent signal processing method based on augmented reality, including:

[0008] S1: Monitor and obtain building vibration or settlement signal data, pre-process it through the bilateral filtering algorithm built into the Internet of Things, transmit it to the edge server for storage, configure the AR device and connect the AR device to the edge server through the Internet of Things;

[0009] S2: construct a signal dynamic compensation model based on RLS filter and particle swarm algorithm, and configure the constructed signal dynamic compensation model to the corresponding edge server, perform dynamic filtering and signal compensation processing on the stored vibration or settlement signal data, and obtain secondary filtering signal data;

[0010] S3: Obtain the corresponding three-dimensional spatial data of the monitored building, and use the edge server to configure the three-dimensional model generation algorithm, build the three-dimensional model of the building using the three-dimensional model generation algorithm according to the acquired three-dimensional spatial data, map the secondary filtering signal data to the constructed three-dimensional model of the building, dynamically adjust the model shape according to the data changes, and simulate the building vibration or settlement process;

[0011] S4: An assessment model is constructed based on a fuzzy algorithm, and the constructed assessment model is built into the edge server. According to the acquired secondary filtering signal data and the building vibration or settlement process, the dynamic risk level of the building vibration or settlement is obtained through the assessment model, and the acquired dynamic risk level is marked on the constructed three-dimensional building model with a color label. At the same time, the risk level is updated in real time using the real-time monitored vibration or settlement signal data;

[0012] S5: The color-labeled 3D building model is transmitted to the configured AR device through the Internet of Things, and the risk level warning conditions configured in the AR device are used to issue a warning according to the change status of the color marked in the 3D building model. According to the warning status, the vibration or settlement risk processing strategy library is automatically called to obtain the corresponding processing strategy, and the obtained strategy is implemented and displayed on the AR device.

[0013] Specifically, the specific steps of constructing the signal dynamic compensation model in S2 include:

[0014] S201, let the vibration or settlement signal data sequence at time t be x t , x t Multiply them by M phase rotation factors respectively to obtain M sub-signal frequency shifts with different phase offsets in, represents the frequency shift of the mth sub-signal with phase shift;

[0015] S202, construct a sub-signal filter compensation array using M filters, and calculate the parameter sequence α corresponding to the M filters in the sub-signal filter compensation array = (α1…α m …α M ) to initialize, and obtain the Input into the corresponding M filters in the sub-signal filter compensation array, and output the compensation signal sequence Among them, α m represents the parameter subsequence of the mth filter among M filters, represents the compensation signal output by the mth filter; the M sub-signal input quantities with different phase offsets correspond one to one with the filters in the sub-signal filter compensation array;

[0016] S203, according to the obtained compensation signal sequence The reconstructed compensation signal data is obtained by using a weighted strategy. The specific formula of the weighted strategy is:

[0017]

[0018] in, represents the reconstructed compensation signal data, h m represents the weight vector of the mth filter in the sub-signal filter compensation array, represents the parameter subsequence after dynamic compensation of the mth filter, represents the weighted compensation coefficient of the parameter subsequence corresponding to the mth filter in the parameter sequence output by the filter parameter adjustment submodel in the signal dynamic compensation model, z -m Represents the signal delay operator.

[0019] Specifically, the specific steps of constructing the signal dynamic compensation model in S2 also include:

[0020] S204, xt 输 The reference signal sub-model constructed by the noise reduction encoder in the signal dynamic compensation model is input to obtain the reference signal sequence r t ;

[0021] S205, by obtaining a reference signal sequence r t and x t Get the signal deviation value e t , the obtained signal deviation value e t Input into the filter parameter adjustment submodel in the signal dynamic compensation model to obtain the weighted compensation coefficient sequence corresponding to the M filter parameter subsequences

[0022] S206: The obtained weighted compensation coefficient sequence Feedback is sent to the corresponding filter in the sub-signal filter compensation array, and the parameter sub-sequences of the M filters are adjusted in real time through the weighted strategy formula of S203;

[0023] S207, setting signal deviation threshold when When , the signal compensation is stopped and the reconstructed compensation signal data is output Otherwise, continue to compensate iteratively until the Stop when.

[0024] Specifically, the specific steps of constructing the filter parameter adjustment sub-model in the signal dynamic compensation model include:

[0025] S2031, using the parameter sequences (α1…α m …α M ) constructs M particles of the particle swarm algorithm, and uses the parameter value corresponding to each filter parameter subsequence to construct the position vector P corresponding to each particle m and the velocity vector V m , where P m Indicates m The position vector of the mth particle is constructed by the filter parameter subsequence, V m represents the velocity vector of the mth particle constructed by the mth filter parameter subsequence;

[0026] S2032. Use the obtained M particles to construct a fitness function, and use the fitness function to evaluate each particle. The fitness function is specifically: where f m Represents the fitness evaluation value of the mth particle;

[0027] S2033, based on the obtained fitness evaluation values ​​of the M particles, calculate and obtain the mean value f of the fitness values ​​corresponding to the M particles a and standard deviation f std , according to the obtained mean f a and standard deviation f std Set the stop judgment condition of the particle swarm algorithm. The stop judgment condition includes:

[0028]

[0029] S2034, the signal deviation value e t Input into the particle swarm algorithm constructed by S2031-S2033, perform an iterative sequence according to the obtained stop judgment condition, and when the stop judgment condition is met, output the weighted compensation coefficients of the corresponding parameter subsequences of the M filters

[0030] Specifically, the dynamic risk level in S4 includes three risk levels. The three risk levels and the corresponding judgment intervals include: t When <40, it corresponds to low risk level. When 40≤g t ≤70, corresponding to medium risk level, when g t >70, corresponding to high risk level, g t It represents the risk score obtained by evaluating the secondary filtering signal data obtained at time t through the evaluation model.

[0031] Specifically, the specific steps of using color labels to mark the acquired dynamic risk levels in the constructed three-dimensional building model in S4 include:

[0032] Green, yellow and red labels are set for low risk level, medium risk level and high risk level respectively. t and discrimination interval, using the color coding algorithm to encode in the same color to obtain a mapping relationship table corresponding to the risk score and color depth;

[0033] The acquired mapping relationship table is configured into the AR device. When the corresponding risk score is input, the corresponding position of the building three-dimensional model is marked with the corresponding color depth through the mapping relationship table.

[0034] The intelligent signal processing system based on augmented reality includes: data preprocessing module, signal processing module, model building and mapping module and risk labeling module;

[0035] The data preprocessing module includes a data monitoring unit and a preprocessing and transmission unit; the data monitoring unit is used to monitor and obtain the original signal data of the vibration or settlement of the building in real time; the preprocessing and transmission unit is used to preprocess the original vibration or settlement signal data using the built-in bilateral filtering algorithm in the Internet of Things, and transmit the preprocessed signal data to the edge server for storage;

[0036] The signal processing module includes a dynamic compensation unit and a model deployment unit; the dynamic compensation unit is used to perform dynamic filtering and signal compensation processing on the stored signal data through the deployed signal dynamic compensation model to obtain secondary filtered signal data.

[0037] Specifically, the model building and mapping module includes a three-dimensional model generation unit and a data mapping unit;

[0038] The three-dimensional model generation unit is used to obtain the three-dimensional spatial data of the corresponding monitored building, and use the three-dimensional model generation algorithm configured by the edge server to construct a three-dimensional model of the building; the data mapping unit is used to map the secondary filtering signal data to the constructed three-dimensional model of the building, and dynamically adjust the model shape according to the data changes to simulate the building vibration or settlement process.

[0039] Specifically, the risk labeling module includes a risk assessment unit and a risk labeling unit; the risk assessment unit is used to obtain the dynamic risk level of building vibration or settlement through an assessment model based on the secondary filtering signal data and the building vibration or settlement process; the risk labeling unit is used to use color labels to mark the obtained dynamic risk level on the constructed three-dimensional building model, and use real-time monitored vibration or settlement signal data to update the risk level in real time.

[0040] Specifically, a computer-readable storage medium stores computer instructions, which, when executed, execute an intelligent signal processing method based on augmented reality.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] In view of the deficiencies in the prior art, the present invention combines the preprocessing of the bilateral filtering algorithm with the efficient data processing capability of the edge server. The method can quickly process a large amount of real-time data transmitted by IoT devices, significantly reduce processing delays, and meet application scenarios with high real-time requirements, especially in high-frequency, high-density vibration or settlement monitoring scenarios, thereby improving the accuracy and response speed of data processing. In addition, the introduction of the signal dynamic compensation model, especially the innovative combination of the RLS filter and the particle swarm algorithm, realizes dynamic filtering and compensation processing of the signal, and can adaptively adjust the filtering parameters according to environmental changes and signal characteristics, thereby improving the adaptability of the processing algorithm to complex environments and high-frequency signals, and effectively solving the limitations of traditional methods in processing specific types of signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 This is a flow chart of an intelligent signal processing method based on augmented reality according to Embodiment 1 of the present invention;

[0044] Figure 2 This is a logical structure diagram of a signal dynamic compensation model according to Embodiment 1 of the present invention;

[0045] Figure 3 This is a system analysis flow chart of an intelligent signal processing method based on augmented reality according to Embodiment 2 of the present invention. DETAILED DESCRIPTION

[0046] Example 1

[0047] See also Figure 1 , an embodiment of the present invention provides: an intelligent signal processing method based on augmented reality, comprising the following steps:

[0048] S1: monitoring and acquiring building vibration or settlement signal data, preprocessing it through the built-in bilateral filtering algorithm in the Internet of Things, transmitting it to the edge server for storage, configuring an AR device and connecting the AR device to the edge server through the Internet of Things; the AR device configured in this embodiment is a smart AR glasses, which can realize the functions of signal reception and processing result display in this embodiment through the Internet of Things; further, in this embodiment, acquiring building vibration or settlement signal data is acquired through vibration monitoring or settlement monitoring; further, the building vibration or settlement signal data acquired in this embodiment includes the location information of the corresponding vibration or settlement point; the built-in bilateral filtering algorithm in the Internet of Things is built-in in the sensor transmitted by the Internet of Things;

[0049] S2: Construct a signal dynamic compensation model based on the RLS filter, and configure the constructed signal dynamic compensation model to the corresponding edge server, perform dynamic filtering and signal compensation processing on the stored vibration or settlement signal data, and obtain secondary filtering signal data; For further information, please refer to Figure 2In this embodiment, the specific steps of constructing the signal dynamic compensation model include:

[0050] S201, let the vibration or settlement signal data sequence at time t be x t , x t Respectively with M phase rotation factors (e -i2πn …e -i2πmn …e -i2πMn ) to obtain the frequency shift of M sub-signals with different phase offsets Where i represents an imaginary number, m represents the index of the filter, and n represents the time index corresponding to each data in the vibration or settlement signal data sequence; represents the frequency shift of the mth sub-signal with phase shift;

[0051] S202, construct a sub-signal filter compensation array using M filters, and calculate the parameter sequence α corresponding to the M filters in the sub-signal filter compensation array = (α1…α m …α M ) to initialize, and obtain the Input into the corresponding M filters in the sub-signal filter compensation array, and output the compensation signal sequence Among them, α m represents the parameter subsequence of the mth filter among M filters, represents the compensation signal output by the mth filter; the M sub-signal input quantities with different phase offsets correspond one to one with the filters in the sub-signal filter compensation array; further, the filters in this embodiment include a low-pass filter, a high-pass filter, a band-pass filter and a Kalman filter;

[0052] S203, according to the obtained compensation signal sequence The reconstructed compensation signal data is obtained by using a weighted strategy. The specific formula of the weighted strategy is:

[0053]

[0054] in, represents the reconstructed compensation signal data, h m represents the weight vector of the mth filter in the sub-signal filter compensation array, represents the parameter subsequence after dynamic compensation of the mth filter, represents the weighted compensation coefficient of the parameter subsequence corresponding to the mth filter in the parameter sequence output by the filter parameter adjustment submodel in the signal dynamic compensation model, z -m represents the signal delay operator. In this embodiment, the signal delay operator z -mIt is used to adjust the order or delay of the outputs of each filter so that they can be correctly combined in the final reconstructed signal. This is because there may be various noises and interferences in the actual system, resulting in slight time differences between the outputs of different filters. In order to obtain accurate results, these outputs need to be properly synchronized.

[0055] S204, x t Input into the reference signal sub-model constructed by the noise reduction encoder in the signal dynamic compensation model to obtain the reference signal sequence r t ;

[0056] S205, by obtaining a reference signal sequence r t and Get the signal deviation value e t , the obtained signal deviation value e t Input into the filter parameter adjustment submodel in the signal dynamic compensation model to obtain the weighted compensation coefficient sequence corresponding to the M filter parameter subsequences

[0057] S206: The obtained weighted compensation coefficient sequence Feedback is sent to the corresponding filter in the sub-signal filter compensation array, and the parameter sub-sequences of the M filters are adjusted in real time through the weighted strategy formula of S203;

[0058] S207, setting signal deviation threshold when When , the signal compensation is stopped and the reconstructed compensation signal data is output Otherwise, continue to compensate iteratively until the Stop when.

[0059] Furthermore, the specific steps of constructing the filter parameter adjustment sub-model in the signal dynamic compensation model include:

[0060] S2031, using the parameter sequences (α1…α m …α M ) constructs M particles of the particle swarm algorithm, and uses the parameter value corresponding to each filter parameter subsequence to construct the position vector P corresponding to each particle m and the velocity vector V m , where P m represents the position vector of the mth particle constructed by the mth filter parameter subsequence, V mrepresents the velocity vector of the mth particle constructed by the mth filter parameter subsequence; in this embodiment, filters are used in one-to-one correspondence with the constructed particles, that is, each particle is constructed by a filter in the sub-signal filter compensation array; further, each filter parameter subsequence includes the following variables: cutoff frequency, quality factor and phase offset;

[0061] S2032. Use the obtained M particles to construct a fitness function, and use the fitness function to evaluate each particle. The fitness function is specifically: where f m Represents the fitness evaluation value of the mth particle;

[0062] S2033, based on the obtained fitness evaluation values ​​of the M particles, calculate and obtain the mean value f of the fitness values ​​corresponding to the M particles a and standard deviation f std , according to the obtained mean f a and standard deviation f std Set the stop judgment condition of the particle swarm algorithm. The stop judgment condition includes:

[0063]

[0064] S2034, the signal deviation value e t Input into the particle swarm algorithm constructed by S2031-S2033, perform an iterative sequence according to the obtained stop judgment condition, and when the stop judgment condition is met, output the weighted compensation coefficients of the corresponding parameter subsequences of the M filters

[0065] This process achieves fine processing of vibration or settlement signals through dynamic phase adjustment and filtering compensation strategies, which not only effectively suppresses noise and improves signal purity, but also ensures the accuracy and timeliness of signal compensation under various environmental conditions through real-time parameter adaptive adjustment and iterative optimization. The final output reconstructed compensation signal data is more accurate and reliable, which significantly enhances the stability and robustness of the signal processing system and provides a high-quality data foundation for subsequent analysis and decision-making.

[0066] S3: Acquire the corresponding three-dimensional spatial data of the monitored building, and use the edge server to configure a three-dimensional model generation algorithm, construct a three-dimensional model of the building using the three-dimensional model generation algorithm according to the acquired three-dimensional spatial data, map the secondary filtering signal data to the constructed three-dimensional model of the building, dynamically adjust the model shape according to data changes, and simulate the vibration or settlement process of the building; further, the specific steps of S3 in this embodiment include:

[0067] S301, using drone aerial photography to obtain three-dimensional spatial coordinate data of the building, and preprocessing the collected data; the three-dimensional spatial coordinate data includes information such as latitude, longitude and height of key points of the building;

[0068] S302, configuring a building information model (BIM) software interface on the edge server, and inputting the acquired three-dimensional spatial coordinate data into the building information model to obtain a three-dimensional building model; further, the three-dimensional construction model configured in the building information model (BIM) software interface is implemented by an existing pre-trained model; wherein the three-dimensional construction model includes a geometric modeling engine, a parametric modeling tool, and an IFC standard compatible interface;

[0069] S303, mapping the secondary filtered signal data to the key area nodes of the building three-dimensional model according to the time series and location information, and using the computational simulation software to obtain the changing state of the vibration or settlement of the building at different time points; the computational simulation software in this step includes ANSYS, SAP2000 / ETABS, ABAQUS and Midas Civil.

[0070] S4: construct an evaluation model based on a fuzzy algorithm, and embed the constructed evaluation model into the edge server. According to the acquired secondary filtering signal data and the building vibration or settlement process, the dynamic risk level of the building vibration or settlement is obtained through the evaluation model, and the acquired dynamic risk level is marked on the constructed three-dimensional building model with a color label. At the same time, the risk level is updated in real time using the real-time monitored vibration or settlement signal data; further, in this embodiment, the evaluation model is constructed using a comprehensive evaluation algorithm; further, in this embodiment, the specific steps of obtaining the dynamic risk level include:

[0071] S401, obtaining an evaluation factor set according to the acquired secondary filtering signal data and the building vibration or settlement process data; the evaluation factor set includes the maximum displacement, frequency, duration, acceleration peak value, etc. during the building vibration or settlement process;

[0072] S402, define a risk level set and a corresponding assessment score interval, wherein the risk level set includes three risk levels, and the three risk levels and the corresponding discrimination intervals include: when g t When <40, it corresponds to low risk level. When 40≤g t ≤70, corresponding to medium risk level, when g t >70, corresponding to high risk level, g t It indicates that the secondary filtering signal data obtained at time t is evaluated by the evaluation model to obtain the risk score;

[0073] S403, using the constructed assessment factor set and risk level set, through the membership function, obtain the membership value matrix of each assessment factor to each risk level, and calculate the weight of each assessment factor through the entropy weight method;

[0074] S404, inputting the obtained assessment factor set, risk level set, assessment score interval, membership value matrix and corresponding weights into the comprehensive fuzzy algorithm to obtain the risk score and risk level of the corresponding location area of ​​the building;

[0075] This process significantly enhances the intelligence and accuracy of building safety assessment through the real-time monitoring system of the edge server integrated with fuzzy algorithm. The system can process the secondary filtering signal data in real time, comprehensively analyze the key indicators of the building during vibration or settlement, such as maximum displacement, frequency, duration and acceleration peak, and then dynamically evaluate the risk status of the building structure. By finely dividing the risk level and optimizing the weight of the assessment factor using the entropy weight method, the model can more accurately reflect the immediate risk level of each area, and intuitively display it on the three-dimensional model with color labels, which is convenient for quickly identifying potential dangerous areas. The real-time update mechanism ensures the timeliness of risk monitoring, allowing managers to take timely measures to effectively prevent building safety accidents, ensure personnel safety and structural stability, and greatly improve the efficiency and reliability of building safety management.

[0076] Furthermore, the specific steps of using color labels to mark the acquired dynamic risk levels in the constructed three-dimensional building model include:

[0077] Green, yellow and red labels are set for low risk level, medium risk level and high risk level respectively. t and discrimination interval, using the color coding algorithm to encode in the same color to obtain a mapping relationship table corresponding to the risk score and color depth;

[0078] The acquired mapping relationship table is configured in the AR device, and when the corresponding risk score is input, the corresponding position of the three-dimensional model of the building is marked with the corresponding color depth through the mapping relationship table.

[0079] Furthermore, the specific steps of obtaining a mapping relationship table corresponding to risk scores and color depths include:

[0080] S405, dividing the discrimination interval of the same color into 5 sub-intervals on average, generating 5 color gradients by adjusting saturation and brightness in the HSV color space through the 5 sub-intervals and corresponding risk scores, and converting the generated color gradients into corresponding RGB values;

[0081] S406. Establish a table or function for saving or mapping the color gradient and the corresponding RGB value obtained in S405. When the risk score is calculated, the output is the corresponding RGB value, and the corresponding area of ​​the building is marked with different color depth levels; further, the color display is displayed in real time through an AR device;

[0082] Exemplarily, taking the assessment score interval corresponding to yellow as an example, the mapping relationship between the risk score and the color depth is: when the risk score is 40-45 points: RGB (255, 255, 153) (light yellow), score 46-51 points: RGB (255, 255, 205) (slightly darker light yellow), score 52-58 points: RGB (255, 255, 0) (pure yellow), score 59-64 points: RGB (255, 165, 0) (orange-yellow), score 65-70 points: RGB (255, 0, 0) (orange gradually transitions to red); for the mapping relationship between the risk score and color depth of green and red, refer to the mapping relationship of yellow, which will not be repeated here. In addition, those skilled in the art can also perform corresponding color annotation and color depth level classification according to other commonly used colors used in this field;

[0083] S407. The building vibration or settlement signal data obtained through real-time monitoring is used to generate corresponding risk scores through the evaluation model. The risk levels of different colors and color depths are displayed on the three-dimensional building model displayed in the AR device through the obtained risk scores and the mapping relationship table, and the corresponding risk scores are displayed on the corresponding colors and color depth levels.

[0084] This process uses a refined color coding algorithm to present the dynamic risk level of building vibration or settlement in real time on the three-dimensional model in the AR device with intuitive and continuous color depth changes. This innovation not only improves the accuracy and efficiency of risk identification, but also enhances the scientific basis for decision-making, allowing managers to quickly grasp the distribution and evolution trends of risks and take effective measures to intervene in time. At the same time, this method is also highly flexible and scalable, supporting the adjustment of color mapping relationships according to actual needs, meeting application requirements in different scenarios, and improving the overall adaptability of signal processing.

[0085] S5: The color-labeled 3D building model is transmitted to the configured AR device through the Internet of Things, and the risk level warning conditions configured in the AR device are used to issue a warning according to the change status of the color marked in the 3D building model. According to the warning status, the vibration or settlement risk processing strategy library is automatically called to obtain the corresponding processing strategy, and the obtained strategy is displayed in real time on the AR device.

[0086] Furthermore, in this embodiment, the risk level warning condition is that when g tWhen the value is ≥40, the risk warning condition is reached, and the risk warning signal is transmitted to the user through the configuration of the Internet of Things;

[0087] In this embodiment, the specific steps of implementing and displaying the acquired strategy on the AR device include:

[0088] S501, input the obtained warning signal and risk score into the vibration or settlement risk processing strategy library, match the corresponding risk processing strategy using reasoning and matching algorithms, and transmit the obtained strategy to the AR glasses device equipped by the user through the Internet of Things for real-time display; further, the steps of constructing the vibration or settlement risk processing strategy library in this embodiment include:

[0089] S5011. Collect and preprocess the vibration or settlement risk treatment methods and strategies to obtain the corresponding text data set;

[0090] S5012. Entity and relationship extraction model constructed by Chinese pre-trained Bert algorithm is used to extract entities and relationships from corresponding text data sets, obtain strategy triples, and input the obtained strategy triples into the graph database to construct a vibration or settlement risk processing strategy library;

[0091] S5013. Configure an inference model at the interface of the vibration or subsidence risk processing strategy library. When the input warning signal and risk score do not match the corresponding processing strategy, call the configured inference model and text or video generation algorithm to infer and generate the corresponding processing strategy based on the existing vibration or subsidence risk processing strategy library and the warning signal and risk score. In this embodiment, the inference models include BART, BERT and GPT; the text or video generation algorithms include Flamingo and BlackEye models.

[0092] Example 2

[0093] See also Figure 3 , another embodiment provided by the present invention: an intelligent signal processing system based on augmented reality, comprising: a data preprocessing module, a signal processing module, a model building and mapping module, a risk labeling module and a display and warning module;

[0094] The data preprocessing module is used to obtain the original signal data of building vibration or settlement and perform preliminary processing; the data preprocessing module includes a data monitoring unit and a preprocessing and transmission unit;

[0095] The data monitoring unit is used to monitor and obtain the original signal data of the vibration or settlement of the building in real time; the preprocessing and transmission unit is used to preprocess the original vibration or settlement signal data using the bilateral filtering algorithm built into the Internet of Things, and transmit the preprocessed signal data to the edge server for storage;

[0096] A signal processing module, used to perform secondary processing on the signal data saved to the edge server using the deployed model to improve the data quality; the signal processing module includes a dynamic compensation unit and a model deployment unit;

[0097] A model deployment unit is used to configure the constructed signal dynamic compensation model to the corresponding edge server; a dynamic compensation unit is used to dynamically filter and perform signal compensation processing on the stored signal data through the deployed signal dynamic compensation model to obtain secondary filtered signal data;

[0098] The model building and mapping module is used to build a three-dimensional model according to the three-dimensional spatial data of the building and map the processed signal data onto the three-dimensional model; the model building and mapping module includes a three-dimensional model generation unit and a data mapping unit;

[0099] A three-dimensional model generation unit is used to obtain the three-dimensional spatial data of the corresponding monitored building, and use the three-dimensional model generation algorithm configured by the edge server to construct a three-dimensional model of the building; a data mapping unit is used to map the secondary filtering signal data to the constructed three-dimensional model of the building, and dynamically adjust the model shape according to the data changes to simulate the vibration or settlement process of the building;

[0100] The risk marking module is used to build an assessment model based on fuzzy algorithms, assess the risk level of building vibration or settlement, and mark it on the three-dimensional model;

[0101] The risk labeling module includes a risk assessment unit and a risk labeling unit; the risk assessment unit is used to obtain the dynamic risk level of building vibration or settlement through the assessment model based on the secondary filtering signal data and the building vibration or settlement process; the risk labeling unit is used to label the obtained dynamic risk level on the constructed three-dimensional building model with color labels, and update the risk level in real time using the real-time monitored vibration or settlement signal data;

[0102] The display and warning module is used to transmit the three-dimensional model with risk annotations to the AR device and to provide warnings and strategy displays according to the risk level; the display and warning module includes a virtual transmission warning unit and a strategy display unit;

[0103] The virtual transmission warning unit is used to transmit the color-labeled building three-dimensional model to the configured AR device through the Internet of Things, and use the risk level warning conditions configured in the AR device to issue a warning based on the change status of the labeled color in the building three-dimensional model; the strategy display unit is used to automatically call the vibration or settlement risk processing strategy library according to the warning status, obtain the corresponding processing strategy, and implement and display it on the AR device.

[0104] Example 3

[0105] A computer-readable storage medium stores computer instructions, which, when executed, execute an intelligent signal processing method based on augmented reality.

[0106] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in the field may also change, modify, replace and modify the above-mentioned embodiments without departing from the scope of protection of the purpose of the present invention and the claims, and all of these are within the protection of the present invention.

[0107] If the disclosed technical solution involves personal information, the product using the disclosed technical solution has clearly informed the personal information processing rules and obtained the individual's voluntary consent before processing the personal information. If the disclosed technical solution involves sensitive personal information, the product using the disclosed technical solution has obtained the individual's separate consent before processing the sensitive personal information, and at the same time meets the "explicit consent" requirement. For example, on personal information collection devices such as cameras, clear and prominent signs are set to inform that the personal information collection scope has been entered and personal information will be collected. If the individual voluntarily enters the collection scope, it is deemed that he or she agrees to the collection of his or her personal information; or on the device that processes personal information, when the personal information processing rules are notified by obvious signs / information, the individual's authorization is obtained through pop-up information or by asking the individual to upload his or her personal information; among them, the personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the type of personal information processed.

Claims

1. An intelligent signal processing method based on augmented reality, characterized in that: include: S1: Monitor and obtain building vibration or settlement signal data, pre-process it through the bilateral filtering algorithm built into the Internet of Things, transmit it to the edge server for storage, configure the AR device and connect the AR device to the edge server through the Internet of Things; S2: construct a signal dynamic compensation model based on RLS filter and particle swarm algorithm, and configure the constructed signal dynamic compensation model to the corresponding edge server, perform dynamic filtering and signal compensation processing on the stored vibration or settlement signal data, and obtain secondary filtering signal data; S3: Obtain the corresponding three-dimensional spatial data of the monitored building, and use the edge server to configure the three-dimensional model generation algorithm, build the three-dimensional model of the building using the three-dimensional model generation algorithm according to the acquired three-dimensional spatial data, map the secondary filtering signal data to the constructed three-dimensional model of the building, dynamically adjust the model shape according to the data changes, and simulate the building vibration or settlement process; S4: An assessment model is constructed based on a fuzzy algorithm, and the constructed assessment model is built into the edge server. According to the acquired secondary filtering signal data and the building vibration or settlement process, the dynamic risk level of the building vibration or settlement is obtained through the assessment model, and the acquired dynamic risk level is marked on the constructed three-dimensional building model with a color label. At the same time, the risk level is updated in real time using the real-time monitored vibration or settlement signal data; S5: The color-labeled 3D building model is transmitted to the configured AR device through the Internet of Things, and the risk level warning conditions configured in the AR device are used to issue a warning according to the change status of the color marked in the 3D building model. According to the warning status, the vibration or settlement risk processing strategy library is automatically called to obtain the corresponding processing strategy, and the obtained strategy is implemented and displayed on the AR device.

2. The intelligent signal processing method based on augmented reality according to claim 1, characterized in that: The specific steps of constructing the signal dynamic compensation model in S2 include: S201, assuming that the vibration or settlement signal data sequence at time t is xt, multiply xt by M phase rotation factors respectively to obtain M sub-signal frequency shifts with different phase offsets in, represents the frequency shift of the mth sub-signal with phase shift; S202, construct a sub-signal filter compensation array using M filters, and calculate the parameter sequence α corresponding to the M filters in the sub-signal filter compensation array = (α1…αm…α M ) to initialize, and obtain the Input into the corresponding M filters in the sub-signal filter compensation array, and output the compensation signal sequence Among them, αm represents the parameter subsequence of the mth filter among M filters, represents the compensation signal output by the mth filter; the M sub-signal input quantities with different phase offsets correspond one-to-one to the filters in the sub-signal filter compensation array; S203, according to the obtained compensation signal sequence The reconstructed compensation signal data is obtained by using a weighted strategy. The specific formula of the weighted strategy is: in, represents the reconstructed compensation signal data, hm represents the weight vector of the mth filter in the sub-signal filter compensation array, represents the parameter subsequence after dynamic compensation of the mth filter, It represents the weighted compensation coefficient of the parameter subsequence corresponding to the mth filter in the parameter sequence output by the filter parameter adjustment submodel in the signal dynamic compensation model, and zm represents the signal delay operator.

3. The intelligent signal processing method based on augmented reality according to claim 2, characterized in that: The specific steps of constructing the signal dynamic compensation model in S2 also include: S204, inputting xt into the reference signal sub-model constructed by the noise reduction encoder in the signal dynamic compensation model to obtain a reference signal sequence rt; S205, obtain the signal deviation value et through the obtained reference signal sequence rt and xt, input the obtained signal deviation value et into the filter parameter adjustment submodel in the signal dynamic compensation model, and obtain the weighted compensation coefficient sequence corresponding to the M filter parameter subsequences S206: The obtained weighted compensation coefficient sequence Feedback is sent to the corresponding filter in the sub-signal filter compensation array, and the parameter sub-sequences of the M filters are adjusted in real time through the weighted strategy formula of S203; S207, setting signal deviation threshold when When , the signal compensation is stopped and the reconstructed compensation signal data is output Otherwise, continue to compensate iteratively until the Stop when.

4. The intelligent signal processing method based on augmented reality according to claim 3, characterized in that: The specific steps of constructing the filter parameter adjustment submodel in the signal dynamic compensation model include: S2031, using the parameter sequence (α1…αm…α M ) constructs M particles of the particle swarm algorithm, and uses the parameter values ​​corresponding to each filter parameter subsequence to construct the position vector Pm and velocity vector Vm corresponding to each particle, where Pm represents the position vector of the mth particle constructed by the mth filter parameter subsequence, and Vm represents the velocity vector of the mth particle constructed by the mth filter parameter subsequence; S2032. Use the obtained M particles to construct a fitness function, and use the fitness function to evaluate each particle. The fitness function is specifically: Where fm represents the fitness evaluation value of the mth particle; S2033, based on the obtained fitness evaluation values ​​of the M particles, calculate the mean fa and standard deviation fstd of the fitness values ​​corresponding to the M particles, and set the stop judgment condition of the particle swarm algorithm according to the obtained mean fa and standard deviation fstd, and the stop judgment condition specifically includes: S2034, input the signal deviation value et into the particle swarm algorithm constructed by S2031-S2033, perform an iterative sequence according to the obtained stop judgment condition, and when the stop judgment condition is met, output the weighted compensation coefficients of the parameter subsequences corresponding to the M filters 5. The intelligent signal processing method based on augmented reality according to claim 4, characterized in that: The dynamic risk level in S4 includes three risk levels, and the three risk levels and corresponding judgment intervals include: when gt<40, it corresponds to a low risk level, when 40≤gt≤70, it corresponds to a medium risk level, and when gt>70, it corresponds to a high risk level, and gt represents the risk score obtained by evaluating the secondary filtering signal data obtained at time t through the evaluation model.

6. The intelligent signal processing method based on augmented reality according to claim 5, characterized in that: The specific steps of using color labels to mark the acquired dynamic risk levels in the constructed three-dimensional building model in S4 include: Green, yellow and red labels are set for the low risk level, medium risk level and high risk level respectively, and a mapping relationship table corresponding to the risk score and the color depth is obtained by encoding in the same color using a color coding algorithm according to the set color labels, risk score gt and discrimination interval; The acquired mapping relationship table is configured in the AR device, and when the corresponding risk score is input, the corresponding position of the three-dimensional model of the building is marked with the corresponding color depth through the mapping relationship table.

7. An intelligent signal processing system based on augmented reality, which is used to implement the intelligent signal processing method based on augmented reality according to any one of claims 1 to 6, characterized in that: include: Data preprocessing module, signal processing module, model building and mapping module, and risk annotation module; The data preprocessing module includes a data monitoring unit and a preprocessing and transmission unit; the data monitoring unit is used to monitor and obtain the original signal data of the vibration or settlement of the building in real time; The preprocessing and transmission unit is used to preprocess the original vibration or settlement signal data using the bilateral filtering algorithm built into the Internet of Things, and transmit the preprocessed signal data to the edge server for storage; The signal processing module includes a dynamic compensation unit and a model deployment unit; the dynamic compensation unit is used to perform dynamic filtering and signal compensation processing on the stored signal data through the deployed signal dynamic compensation model to obtain secondary filtered signal data.

8. The intelligent signal processing system based on augmented reality according to claim 7, characterized in that: The model building and mapping module includes a three-dimensional model generation unit and a data mapping unit; The three-dimensional model generation unit is used to obtain the three-dimensional spatial data of the corresponding monitored building, and use the three-dimensional model generation algorithm configured by the edge server to construct a three-dimensional model of the building; the data mapping unit is used to map the secondary filtering signal data to the constructed three-dimensional model of the building, and dynamically adjust the model shape according to the data changes to simulate the building vibration or settlement process.

9. The intelligent signal processing system based on augmented reality according to claim 8, characterized in that: The risk labeling module includes a risk assessment unit and a risk labeling unit; the risk assessment unit is used to obtain the dynamic risk level of building vibration or settlement through an assessment model based on the secondary filtering signal data and the building vibration or settlement process; The risk marking unit is used to mark the acquired dynamic risk level on the constructed three-dimensional building model using color labels, and to update the risk level in real time using the vibration or settlement signal data monitored in real time.

10. A computer-readable storage medium, characterized in that: Computer instructions are stored thereon, and when the computer instructions are executed, the intelligent signal processing method based on augmented reality described in any one of claims 1 to 6 is executed.

Images