Integrated intelligent air flow meter

Abstract

The invention relates to the technical field of air flow meters, and particularly discloses an integrated intelligent air flow meter which performs air detection on the basis of the ultrasonic principle, a pair of ultrasonic transducers is arranged in a sewage deodorization pipeline, and the ultrasonic wave propagation speed is adjusted according to the influence of air flow on the ultrasonic wave propagation speed. And the air flow is preliminarily calculated in combination with pipeline parameters. And then, in order to correct the influence of the temperature change on the ultrasonic propagation speed, the integrated intelligent air flow meter further performs real-time monitoring and data analysis on the temperature in the pipeline, and performs compensation correction on the preliminarily calculated air flow based on a time sequence change mode of the temperature, so that more accurate air flow measurement is realized. In this way, measurement errors caused by temperature fluctuation can be effectively reduced, and the measurement precision and stability of the air flow are improved.

Description

Technical Field

[0001] The present application relates to the technical field of air flow meters, and more specifically, to an integrated intelligent air flow meter. Background Art

[0002] With the acceleration of urbanization and the improvement of environmental protection awareness, it has become particularly important for municipal sewage treatment facilities to control and manage the odorous gases produced while treating sewage. The odorous gases produced by municipal sewage not only affect the quality of life of surrounding residents, but may also pollute the environment. Therefore, establishing an effective sewage deodorization system has become one of the key links in the design and operation of municipal sewage treatment facilities.

[0003] In the process of sewage deodorization, accurate measurement of air flow is an important factor in ensuring the deodorization effect. Air flow directly affects the dilution and diffusion efficiency of odorous substances, as well as the working performance of subsequent treatment equipment (such as biological filters, chemical scrubbers, etc.). Traditional air flow measurement methods mainly rely on differential pressure flowmeters, thermal flowmeters, vortex flowmeters, etc. However, these sensors generally have problems such as complex installation, susceptibility to contamination, high maintenance costs, and strict requirements on flow fields. For example, differential pressure flowmeters are easily affected by the deposition of dirt in the pipeline, resulting in reduced measurement accuracy; thermal flowmeters may be damaged due to high humidity and the presence of corrosive gases in the pipeline; vortex flowmeters do not have ideal measurement results under low speed or turbulent conditions.

[0004] In recent years, with the development of sensor technology and signal processing technology, the flow measurement technology based on the ultrasonic principle has been gradually applied to fluid measurement under various complex working conditions due to its advantages of non-contact, fast response speed, large range ratio, and no medium contamination. Ultrasonic flowmeters determine the flow rate by measuring the time difference or frequency difference of ultrasonic waves propagating in the fluid, and then calculate the flow rate. It is not affected by the fluid medium and is particularly suitable for complex environments such as sewage deodorization pipelines containing corrosive gases and suspended matter.

[0005] However, in practical applications, temperature changes will affect the propagation speed of ultrasonic waves in the air, resulting in measurement errors. Therefore, in an environment where there may be large temperature changes, such as sewage deodorization pipes, how to effectively eliminate the influence of temperature on the measurement results of ultrasonic flow meters has become one of the key issues in improving measurement accuracy.

[0006] Therefore, an optimized integrated intelligent air flow meter is expected. Summary of the invention

[0007] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides an integrated intelligent air flow meter, which performs air detection based on the ultrasonic principle, and deploys a pair of ultrasonic transducers in the sewage deodorization pipeline, and performs a preliminary calculation of the air flow according to the influence of air flow on the ultrasonic propagation velocity and in combination with the pipeline parameters. Then, in order to correct the influence of temperature changes on the ultrasonic propagation velocity, the integrated intelligent air flow meter further monitors and analyzes the temperature in the pipeline in real time, and compensates and corrects the preliminarily calculated air flow based on the time series change pattern of the temperature to achieve more accurate air flow measurement. In this way, the measurement error caused by temperature fluctuations can be effectively reduced, and the measurement accuracy and stability of the air flow can be improved.

[0008] Accordingly, according to one aspect of the present application, an integrated intelligent air flow meter is provided, comprising:

[0009] a first ultrasonic transducer and a second ultrasonic transducer, the first ultrasonic transducer being used to transmit a first ultrasonic detection signal to the second ultrasonic transducer, the second ultrasonic transducer being used to transmit a second ultrasonic detection signal to the first ultrasonic transducer, the first ultrasonic transducer being further used to receive the second ultrasonic detection signal and the second ultrasonic transducer being further used to receive the first ultrasonic detection signal;

[0010] A temperature acquisition unit, used to acquire the time queue of the temperature in the pipeline through a temperature sensor;

[0011] A signal processing unit, configured to determine an initial air flow value based on a first time from the transmission to the reception of the first ultrasonic detection signal and a second time from the transmission to the reception of the second ultrasonic detection signal;

[0012] A signal transmission unit, used for transmitting the time queue of the temperature in the pipeline to the signal processing unit;

[0013] The signal processing unit is further used to perform temperature compensation on the initial air flow value based on the time sequence of the temperature in the pipeline to obtain a corrected air flow value.

[0014] Compared with the prior art, the integrated intelligent air flow meter provided by the present application performs air detection based on the ultrasonic principle. By deploying a pair of ultrasonic transducers in the sewage deodorization pipeline, the air flow is preliminarily calculated based on the influence of air flow on the ultrasonic propagation velocity and the pipeline parameters. Then, in order to correct the influence of temperature changes on the ultrasonic propagation velocity, the integrated intelligent air flow meter further monitors and analyzes the temperature in the pipeline in real time, and compensates and corrects the preliminarily calculated air flow based on the time series change pattern of the temperature to achieve more accurate air flow measurement. In this way, the measurement error caused by temperature fluctuations can be effectively reduced, and the measurement accuracy and stability of air flow can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0016] Figure 1 4 is a block diagram of a signal processing unit in an integrated intelligent air flow meter according to an embodiment of the present application.

[0017] Figure 2 Schematic diagram of data flow of a signal processing unit in an integrated intelligent air flow meter according to an embodiment of the present application.

[0018] Figure 3 It is a block diagram of a fast semantic query subunit in an integrated intelligent air flow meter according to an embodiment of the present application.

[0019] Figure 4 It is a structural block diagram of an integrated intelligent air flow meter according to an embodiment of the present application. DETAILED DESCRIPTION

[0020] Below, the embodiments of the present application will be described in more detail in conjunction with the accompanying drawings, and the above and other purposes, features and advantages of the present application will become more apparent. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the example embodiments described herein.

[0021] In response to the technical problems described in the above background technology, this application proposes an integrated intelligent air flow meter, which performs air detection based on the ultrasonic principle. By deploying a pair of ultrasonic transducers in the sewage deodorization pipeline, the air flow is preliminarily calculated based on the influence of air flow on the ultrasonic propagation velocity and the pipeline parameters. Then, in order to correct the influence of temperature changes on the ultrasonic propagation velocity, the integrated intelligent air flow meter further monitors and analyzes the temperature in the pipeline in real time, and compensates and corrects the preliminarily calculated air flow based on the time series change pattern of the temperature to achieve more accurate air flow measurement. In this way, the measurement error caused by temperature fluctuations can be effectively reduced, and the measurement accuracy and stability of the air flow can be improved.

[0022] Specifically, if Figure 4 As shown, the integrated intelligent air flow meter includes: a first ultrasonic transducer and a second ultrasonic transducer, the first ultrasonic transducer is used to transmit a first ultrasonic detection signal to the second ultrasonic transducer, the second ultrasonic transducer is used to transmit a second ultrasonic detection signal to the first ultrasonic transducer, the first ultrasonic transducer is also used to receive the second ultrasonic detection signal and the second ultrasonic transducer is also used to receive the first ultrasonic detection signal; a temperature acquisition unit, used to acquire a time queue of the temperature in the pipeline through a temperature sensor; a signal processing unit, used to determine an initial air flow value based on a first time from the emission to the reception of the first ultrasonic detection signal and a second time from the emission to the reception of the second ultrasonic detection signal; a signal transmission unit, used to transmit the time queue of the temperature in the pipeline to the signal processing unit; the signal processing unit is also used to perform temperature compensation on the initial air flow value based on the time queue of the temperature in the pipeline to obtain a corrected air flow value.

[0023] Specifically, the integrated intelligent air flow meter described in the present application measures air flow based on the physical phenomenon that the propagation speed of ultrasonic waves in a fluid is affected by the flow direction of the fluid. When the ultrasonic wave propagates along the direction of air flow (downstream), its propagation speed will increase; conversely, when the ultrasonic wave propagates against the direction of air flow (countercurrent), its propagation speed will slow down. Therefore, the present application measures air flow by installing a first ultrasonic transducer and a second ultrasonic transducer inside the pipeline. In order to ensure that the ultrasonic signal can be effectively propagated in the air and accurately received, the deployment method of the ultrasonic transducer and its related precautions will be described in detail below.

[0024] First of all, the selection and preparation of ultrasonic transducers are the basis of the entire deployment process. Selecting a suitable ultrasonic transducer requires consideration of multiple factors, including operating frequency, power output, sensitivity, operating temperature range, etc. The operating frequency determines the propagation characteristics of ultrasound in the fluid, while the power output affects the signal strength and transmission distance. Sensitivity determines the ability of the transducer to receive weak signals, which is especially important in low flow rates or high noise environments. The operating temperature range is a key parameter to ensure that the transducer works properly under different temperature conditions.

[0025] Next, choosing a suitable installation location is crucial. Ideally, the ultrasonic transducer should be installed on a straight pipe section, away from elbows, valves, and other equipment that may cause turbulence. This is because turbulence can interfere with the propagation of ultrasonic signals, resulting in measurement errors. In addition, the installation location should try to avoid the absorption and scattering of signals by bubbles and solid particles in the fluid, which will affect the accuracy of the measurement. This is especially important for environments such as sewage deodorization pipelines that contain a large amount of suspended matter and corrosive gases.

[0026] After determining the installation location, it is necessary to consider the specific installation method of the ultrasonic transducer. Usually, the transducer can be fixed to the outer wall of the pipe by clamping, welding or flange connection. Clamping is the most common installation method, which does not require permanent changes to the pipe and is easy to disassemble and maintain. Welding is suitable for high temperature and high pressure environments, but it will cause permanent changes to the pipe and is not recommended for occasions that require frequent replacement or adjustment. Flange connection is suitable for large-diameter pipes and can provide better sealing and stability.

[0027] During installation, it is also necessary to ensure that the distance and angle between the transducers meet the design requirements. A distance that is too short will result in insufficient signal strength, while a distance that is too long may increase the risk of signal attenuation and noise interference. Generally speaking, the distance between transducers should be several times the diameter of the pipe, and the specific value depends on the measurement requirements. In terms of angle, the transducers are usually installed along the axis of the pipe, but sometimes they are fine-tuned according to the flow field characteristics to optimize the signal propagation path.

[0028] In addition to the physical location and mounting method, the alignment of the transducer is also very important. The transmitter and receiver must be precisely aligned to ensure that the ultrasonic signal can propagate along a straight path. This usually requires the use of special alignment tools and methods, such as laser alignment instruments or special software. During the alignment process, the flow direction and velocity distribution of the fluid in the pipeline must also be considered to ensure that the signal path is consistent with the mainstream direction and reduce the influence of refraction and reflection.

[0029] Specifically, two-way communication is achieved between the first ultrasonic transducer and the second ultrasonic transducer. The first ultrasonic transducer transmits a first ultrasonic detection signal to the second ultrasonic transducer, and the time from the transmission to the reception of the first ultrasonic detection signal is taken as the first time; and the second ultrasonic transducer transmits a second ultrasonic detection signal to the first ultrasonic transducer, and the time from the transmission to the reception of the second ultrasonic detection signal is taken as the second time. The signal processing unit can calculate the average flow velocity of the air based on the difference between the first time and the second time, combined with the diameter of the pipeline, the distance between the first ultrasonic transducer and the second ultrasonic transducer, and the angle between the ultrasonic path and the fluid flow direction. Furthermore, based on the cross-sectional area of ​​the pipeline and the average flow velocity of the air, the initial air flow value can be calculated.

[0030] In a specific example of the present application, the signal processing unit includes: a time difference calculation subunit, which is used to calculate the time difference between the first time and the second time to obtain the downstream-reverse time difference; an air average flow rate calculation subunit, which is used to calculate the air average flow rate value based on the downstream-reverse time difference; an initial air flow calculation subunit, which is used to calculate the product of the pipe cross-sectional area and the air average flow rate value to obtain the initial air flow value. More specifically, the air average flow rate calculation subunit is used to: calculate the air average flow rate value based on the downstream-reverse time difference using the following flow rate calculation formula, wherein the flow rate calculation formula is:

[0031]

[0032] Wherein, L is the diameter of the pipe, d is the distance between the first ultrasonic transducer and the second ultrasonic transducer, θ is the angle between the ultrasonic path and the air flow direction, Δt is the downstream-reverse flow time difference, and v is the average air flow velocity value.

[0033] In particular, considering that temperature changes will affect the propagation speed of ultrasonic waves, thereby causing measurement errors. Therefore, the integrated intelligent air flow meter described in the present application further integrates a temperature acquisition unit for real-time monitoring of temperature changes in the pipeline, and collects the time queue of the temperature in the pipeline through a temperature sensor. Then, the collected time queue of the temperature in the pipeline is transmitted to the signal processing unit through the signal transmission unit for further data processing, and the initial air flow value is temperature compensated and corrected, so as to reduce the impact of temperature fluctuations on the measurement results, thereby improving the accuracy of air flow measurement.

[0034] In this process, the installation of temperature sensors is also an important part of the deployment process. The temperature sensor should be installed close to the ultrasonic transducer to monitor the temperature changes in the pipeline in real time. The temperature sensor can be embedded in the pipeline wall or fixed to the outer wall by clamping or welding. Regardless of the installation method, it is necessary to ensure that the temperature sensor is in full contact with the air to obtain accurate temperature data. In addition, the response time and accuracy of the temperature sensor must also meet the system requirements to ensure the effectiveness of the temperature compensation algorithm.

[0035] After completing the hardware installation, debugging and calibration are required. During the debugging process, you should first check whether all connections are firm and reliable, and whether the power and signal lines are correctly connected. Then, verify the measurement accuracy and stability of the flow meter through testing under standard flow sources or known flow conditions.

[0036] Finally, in order to ensure the long-term stable operation of the flow meter, a detailed maintenance plan needs to be formulated. Regularly check the working status of the ultrasonic transducer and temperature sensor, clean the sediment and dirt in the pipeline, and replace damaged or aging parts.

[0037] At the same time, in order to ensure the reliability of data transmission, the design and implementation of the signal transmission unit needs to consider multiple factors, including data transmission mode, transmission protocol, anti-interference measures, data synchronization mechanism, and fault diagnosis and recovery mechanism. The following is a detailed description of the implementation method of the signal transmission unit:

[0038] Data transmission mode: Depending on the application scenario, you can choose wired transmission or wireless transmission. Wired transmission usually uses cables or optical fibers for data transmission, which has the advantages of long transmission distance, high transmission rate, and strong anti-interference ability. It is suitable for occasions with high requirements for real-time and reliability of data transmission, such as long-distance pipeline monitoring. Wireless transmission uses radio waves or other wireless communication technologies (such as Bluetooth, Wi-Fi, LoRa, etc.) to transmit data, which has the advantages of flexible installation, simple wiring, and easy maintenance. It is suitable for application scenarios with limited space or strong mobility, such as small pipeline monitoring and portable equipment.

[0039] Transmission protocol: Whether it is wired transmission or wireless transmission, you need to choose a suitable transmission protocol to ensure the correct transmission and parsing of data. Commonly used transmission protocols include Modbus, TCP / IP, CAN, RS-485, etc. Modbus is a serial communication protocol widely used in industrial control systems. It supports point-to-point or multi-point communication and is suitable for simple data exchange needs. The TCP / IP protocol is suitable for complex network environments, supports long-distance, high-bandwidth data transmission, and is suitable for distributed systems and remote monitoring. The CAN (Controller Area Network) protocol has high reliability and real-time performance, and is suitable for high-speed data transmission and multi-node communication. The RS-485 protocol is a half-duplex asynchronous communication protocol that supports multi-point communication and is suitable for long-distance, low-rate data transmission. It is widely used in data exchange between industrial field equipment.

[0040] Anti-interference measures: In industrial environments, electromagnetic interference and noise are the main factors affecting the reliability of data transmission. In order to ensure the stable operation of the signal transmission unit, a series of anti-interference measures need to be taken. First, choosing shielded cables or optical fibers for data transmission can effectively reduce the impact of electromagnetic interference. Secondly, using differential signal transmission to cancel out the positive and negative signals can improve the signal's anti-noise ability. In addition, filters and voltage stabilizers can be installed on the signal transmission line to filter out high-frequency noise and voltage fluctuations to ensure the purity and stability of the signal. For wireless transmission methods, the transmission quality and anti-interference ability of wireless signals can be improved by increasing antenna gain, optimizing transmission frequency and power, and adopting frequency hopping technology.

[0041] Data synchronization mechanism: In order to ensure the time consistency between temperature data and ultrasonic signals, the signal transmission unit needs to have data synchronization function. The data synchronization mechanism can be implemented through hardware or software. Hardware synchronization usually uses a synchronized clock or trigger signal to ensure that the temperature sensor and ultrasonic transducer collect data at the same time point. Software synchronization uses a timestamp to mark the acquisition time of each data packet, and the signal processing unit sorts and processes the data according to the timestamp. In order to improve the synchronization accuracy, a high-precision RTC (Real-Time Clock) module can be used to ensure that the time base of each sensor and processing unit is consistent. In addition, cross-device time synchronization can be achieved through the Network Time Protocol (NTP) or GPS synchronization technology, which is suitable for distributed systems and remote monitoring scenarios.

[0042] Fault diagnosis and recovery mechanism: In order to ensure the high availability and reliability of the system, the signal transmission unit needs to have a complete fault diagnosis and recovery mechanism. Fault diagnosis mainly includes hardware fault detection, communication fault detection and data integrity detection. Hardware fault detection detects and reports hardware faults in a timely manner by monitoring the circuit status and sensor working conditions. Communication fault detection detects the status of the communication link through heartbeat signals and timeout mechanisms. Once the communication is interrupted, it immediately starts reconnection or switches to the backup channel. Data integrity detection calculates the checksum of the transmitted data to ensure that no errors occur during the data transmission process. If a data error is detected, the data packet can be resent through the retransmission mechanism until the correct confirmation response is received. The recovery mechanism includes automatic restart, backup channel switching and manual intervention. The automatic restart function can automatically restart the signal transmission unit and restore normal working status when a system abnormality is detected. The backup channel switching function automatically switches to the backup channel when the main channel fails to ensure the continuity of data transmission. Manual intervention allows users to intervene and troubleshoot the system through remote control or on-site operation when a fault that cannot be automatically recovered occurs.

[0043] In summary, in the process of implementing the signal transmission unit, certain steps and precautions need to be followed. First, select the appropriate transmission method and transmission protocol according to the application scenario. For long-distance and high-bandwidth applications, it is recommended to use wired transmission and TCP / IP protocol; for flexible installation and high mobility, it is recommended to use wireless transmission and LoRa protocol. Secondly, design reasonable circuits and communication interfaces to ensure the stability and reliability of signal transmission. For wired transmission, select high-quality shielded cables and connectors to ensure the purity and stability of the signal; for wireless transmission, select high-gain antennas and optimize the transmission frequency to improve signal quality.

[0044] Figure 1 4 is a block diagram of a signal processing unit in an integrated intelligent air flow meter according to an embodiment of the present application. Figure 2 FIG. 1 is a schematic diagram of data flow in a signal processing unit in an integrated intelligent air flow meter according to an embodiment of the present application. Figure 1 and Figure 2As shown, the signal processing unit 100 of the integrated intelligent air flow meter according to the embodiment of the present application includes: a phase space reconstruction subunit 110, which is used to perform phase space reconstruction on the time queue of the temperature in the pipeline to obtain a set of temperature time series subsequences in the pipeline; a time series feature extraction subunit 120, which is used to respectively extract the time series features of each temperature time series subsequence in the pipeline in the set of temperature time series subsequences in the pipeline to obtain a sequence of local time series implicit coding feature vectors of the temperature in the pipeline; an embedded coding subunit 130, which is used to embed the initial air flow value to obtain an initial air flow embedded coding vector; a fast semantic query subunit 140, which is used to perform fast scanning semantic query coding on the sequence of local time series implicit coding feature vectors of the temperature in the pipeline and the initial air flow embedded coding vector to obtain an air flow-temperature in the pipeline query response semantic coding vector; an air flow correction subunit 150, which is used to generate the corrected air flow value based on the air flow-temperature in the pipeline query response semantic coding vector.

[0045] In the above-mentioned signal processing unit 100, the phase space reconstruction subunit 110 is used to perform phase space reconstruction on the time queue of the temperature in the pipeline to obtain a set of time series subsequences of the temperature in the pipeline. It should be understood that the propagation speed of the ultrasonic wave follows a linear relationship with the change of temperature. Therefore, in the temperature compensation correction process of the initial air flow value, it is first necessary to understand the time series dynamic change pattern of the temperature. Considering that temperature changes are usually nonlinear, in order to more accurately capture the dynamic characteristics of temperature changes, the present application uses phase space reconstruction technology to process the time queue of the temperature in the pipeline to convert the one-dimensional temperature time series data into a multi-dimensional phase space to obtain a set of time series subsequences of the temperature in the pipeline. It should be known that the phase space reconstruction technology reconstructs the phase space of temperature changes through a delayed embedding method, which can effectively restore the multi-dimensional dynamic characteristics of the temperature data, thereby revealing the dynamic behavior originally hidden in the temperature time series, which helps to more accurately understand the time series trend and pattern of temperature changes.

[0046] In the specific implementation of phase space reconstruction, first, a time delay parameter τ and an embedding dimension m are selected to construct a phase space subsequence, that is, the pipeline temperature time series subsequence. Each pipeline temperature time series subsequence is in the form of:

[0047] X(i)={T(i),T(i-τ),T(i-2τ),…,T(i-(m-1)τ)}

[0048] Wherein, T(i) is the temperature value at the i-th time point in the time queue of the temperature in the pipeline, X(i) represents the time series subsequence of the temperature in the pipeline, i=(m-1)τ+1,…,N. Wherein, N is the length of the time queue of the temperature in the pipeline, and N-(m-1)τ is the number of constructed time series subsequences of the temperature in the pipeline.

[0049] In the above-mentioned signal processing unit 100, the timing feature extraction subunit 120 is used to respectively extract the timing features of each in-pipe temperature timing subsequence in the set of in-pipe temperature timing subsequences to obtain a sequence of implicitly encoded feature vectors of the local timing of the in-pipe temperature. In a specific example of the present application, each in-pipe temperature timing subsequence in the set of in-pipe temperature timing subsequences is respectively input into a sequence encoder based on an RNN model to obtain a sequence of implicitly encoded feature vectors of the local timing of the in-pipe temperature. That is, in order to effectively capture the dynamic change characteristics of the temperature timing contained in each in-pipe temperature timing subsequence, the present application further adopts an RNN model with excellent performance in time series analysis tasks as a sequence encoder to perform timing feature extraction on each in-pipe temperature timing subsequence in the set of in-pipe temperature timing subsequences. Those skilled in the art should know that the RNN model can effectively memorize dependencies in sequence data through its internal recurrent connection structure, and transfer state information at previous moments to the current moment, thereby capturing the dynamic change characteristics in the temperature time series subsequences in each pipeline, mining the local change patterns of temperature on different time scales, and obtaining a sequence of implicitly encoded feature vectors of the local time series of temperature in the pipeline.

[0050] In the above-mentioned signal processing unit 100, the embedded coding subunit 130 is used to embed the initial air flow value to obtain the initial air flow embedded coding vector. That is, in order to compensate and correct the initial air flow value using the time series change information of the temperature, it is necessary to further convert the initial air flow value and the sequence of the local time series implicit coding feature vector of the temperature in the pipeline into the same feature space. Therefore, the present application adopts the embedded coding technology to process the initial air flow value, and by constructing a mapping function, the initial air flow value is mapped to a high-dimensional feature space to generate an initial air flow embedded coding vector, so as to use a deep learning model to learn the correlation between temperature change and air flow, thereby realizing temperature compensation correction of the initial air flow value. In a specific embodiment of the present application, a fully connected layer can be used as an embedded encoder, and the initial air flow value can be converted into a high-dimensional feature representation that matches the local time series implicit coding feature vector of the temperature in the pipeline by setting a weight matrix.

[0051] In the above-mentioned signal processing unit 100, the fast semantic query subunit 140 is used to perform fast scanning semantic query encoding on the sequence of implicitly encoded feature vectors of the local temporal sequence of the temperature in the pipeline and the embedded encoding vector of the initial air flow to obtain an air flow-temperature query response semantic encoding vector in the pipeline. In the technical solution of the present application, in order to query the temperature change information significantly associated with the initial air flow value from the sequence of implicitly encoded feature vectors of the local temporal sequence of the temperature in the pipeline, so as to enhance the sensitivity of the model to the temperature change information, thereby achieving more efficient temperature compensation correction, the present application proposes a fast scanning semantic query encoding method, which quickly locates the temperature change features most relevant to the air flow by parallelly calculating the correlation between the implicitly encoded feature vectors of the local temporal sequence of the temperature in each pipeline and the embedded encoding vector of the initial air flow, thereby achieving a fast response to temperature changes and improving the data processing speed.

[0052] Figure 3 FIG. 1 is a block diagram of a fast semantic query subunit in an integrated intelligent air flow meter according to an embodiment of the present application. Figure 3 As shown, the fast semantic query subunit 140 includes: a fast matching secondary subunit 141, which is used to filter a subsequence of the local time series implicit coding feature vector of the temperature in the pipe from the sequence of the local time series implicit coding feature vector of the temperature in the pipe based on the semantic correlation between the initial air flow embedded coding vector and each local time series implicit coding feature vector of the temperature in the pipe in the sequence of the local time series implicit coding feature vector of the temperature in the pipe; and a cross-domain query response analysis secondary subunit 142, which is used to perform cross-domain query response analysis on the subsequence of the initial air flow embedded coding vector and the local time series implicit coding feature vector of the temperature in the pipe to obtain the air flow-temperature in the pipe query response semantic coding vector.

[0053] Specifically, the fast matching secondary subunit 141 is used to: first, calculate the mutual information between the initial air flow embedded coding vector and each local temporal implicit coding feature vector of the in-pipe temperature in the sequence of the local temporal implicit coding feature vector of the in-pipe temperature to obtain a sequence of air flow-in-pipe temperature fast semantic matching factors, which is expressed as follows:

[0054] v2={v 21 ,v 22 ,...,v 2i ,...,v 2n}

[0055] x1 = sigmoid(v1)

[0056] x 2i =sigmoid(v2i)

[0057] MI i =H(x1)+H(x 2i )-H(x1,x 2i )

[0058] Wherein, v1 represents the embedded coding vector of the initial air flow, v2 represents the sequence of implicit coding feature vectors of the local temporal sequence of the temperature in the pipeline, and v 21 、v 22 、v 2i and v 2n represents the first, second, i-th and n-th implicit coding feature vectors of the local time series of the temperature in the pipeline in the sequence of the implicit coding feature vectors of the local time series of the temperature in the pipeline, n is the number of the implicit coding feature vectors of the local time series of the temperature in the pipeline, sigmoid represents the sigmoid normalization function, x1 represents the normalized initial air flow embedded coding vector, x 2i represents the i-th normalized initial air flow embedded coding vector, H(x1) represents the information entropy of the normalized initial air flow embedded coding vector, H(x 2i ) represents the information entropy of the i-th normalized initial air flow embedded coding vector, H(x1,x 2i ) represents the joint entropy between the normalized initial air flow embedding code vector and the i-th normalized initial air flow embedding code vector, MI i Represents the i-th air flow-duct temperature fast semantic matching factor.

[0059] Here, the mutual information between the implicitly encoded feature vector of the local time series of the temperature in each pipeline and the embedded encoding vector of the initial air flow is calculated to evaluate the feature correlation between the two. It should be understood that the larger the mutual information value, the more highly correlated the two are; conversely, the smaller the mutual information value, the weaker the correlation between the two. In this way, not only can the temperature features most related to the air flow be effectively identified, but also an objective standard can be provided for subsequent rapid matching in a quantitative manner.

[0060] Next, the first maximum value and the second maximum value are identified from the sequence of the air flow-duct temperature fast semantic matching factors, which can be expressed as:

[0061]

[0062] Among them, max(·) represents the maximum value function, MI max1 and MI max2 Represent the first maximum value and the second maximum value respectively.

[0063] Then, based on the positions of the first maximum value and the second maximum value in the sequence of the air flow-duct temperature fast semantic matching factors, a subsequence of the duct temperature local temporal implicit coding feature vectors for fast matching is determined from the sequence of the duct temperature local temporal implicit coding feature vectors, which is expressed by the formula:

[0064]

[0065] {k sub}={v 2argmax1 ,...,v 2arg max2}

[0066] Wherein, arg max(·) represents the index corresponding to the maximum value, arg max1 and arg max2 represent the position indexes of the first maximum value and the second maximum value in the sequence of the air flow-duct temperature fast semantic matching factor, respectively, and v 2arg max1 and v 2arg max2 represent the implicit encoding feature vectors of the local time series of the temperature in the pipeline corresponding to the first maximum value and the second maximum value, respectively, sub A subsequence of implicitly encoded feature vectors representing the local time series of the temperature in the pipeline.

[0067] That is, by querying the two maximum values ​​from the sequence of air flow-temperature in-duct fast semantic matching factors, and based on the positions of the two maximum values ​​in the sequence, a subsequence of temperature features that are significantly correlated with the initial air flow value is extracted from the sequence of the corresponding local temporal implicit encoding feature vectors of the temperature in the duct, thereby achieving preliminary feature selection and filtering, so as to concentrate resources on processing the most relevant interactive information, reduce unnecessary calculations, reduce noise interference, and improve the generalization ability of the model.

[0068] Specifically, the cross-domain query response analysis secondary subunit 142 is used to: perform a linear transformation on the initial air flow embedded coding vector to obtain a query vector and a value vector, and use a subsequence of the local temporal implicit coding feature vector of the temperature in the pipeline as a subsequence of the key vector, and input the query vector, the value vector and the subsequence of the key vector into a cross-domain query coding module based on a converter structure to obtain the air flow-pipeline temperature query response semantic coding vector, which is expressed as follows:

[0069] v q =v1W q +b q

[0070] v v =v1W v +b v

[0071]

[0072] Among them, W q and W v denote the query embedding matrix and the value embedding matrix respectively, and b q and b v Represent the query bias vector and value bias vector respectively, v q and v v denote the query vector and value vector respectively, (·) T represents the transpose of a vector, v 2j represents the jth local time series implicit coding feature vector of the temperature in the pipeline in the sequence of the local time series implicit coding feature vectors, softmax(·) represents the normalized exponential function, represents a matrix multiplication operation, and V represents the semantic encoding vector of the air flow-temperature in the duct query response.

[0073] That is, a query vector and a value vector are constructed based on the initial air flow embedded coding vector, and a subsequence of the selected in-duct temperature local time series implicit coding feature vector is used as a subsequence of the key vector. Cross-domain query interaction is performed through the converter, and the self-attention mechanism is used to learn the long-distance dependency relationship between temperature information and air flow information, and an air flow-in-duct temperature query response semantic coding vector is generated, thereby realizing a deep interactive fusion of temperature features and air flow features.

[0074] In the above-mentioned signal processing unit 100, the air flow correction subunit 150 is used to generate the corrected air flow value based on the semantic coding vector of the air flow-temperature query response in the pipeline. In a specific example of the present application, the semantic coding vector of the air flow-temperature query response in the pipeline is input into the temperature compensation optimization module based on the decoder to obtain the corrected air flow value. Specifically, the decoder adopts a multi-layer perceptron structure, and comprehensively considers the initial measurement value of the air flow and the time series change pattern of the temperature by performing multi-level feature perception and nonlinear transformation on the semantic coding vector of the air flow-temperature query response in the pipeline, and decodes a more accurate corrected air flow value. During the training process of the decoder, it adopts the back propagation algorithm and the gradient descent optimization strategy to continuously adjust the weights and bias parameters inside the decoder to minimize the error between the predicted corrected air flow value and the true air flow value. Through continuous iterative optimization, the decoder can learn the complex mapping relationship between temperature change and air flow, thereby improving the accuracy of temperature compensation correction.

[0075] In particular, when the sequence of the local temporal implicit coding feature vectors of the temperature in the pipe and the initial air flow embedded coding vector represent the temporal distribution of the local time domain implicit correlation characteristics of the temperature in the pipe and the initial air flow embedded coding characteristics respectively, when calculating the semantic query coding therebetween, the air flow-temperature in the pipe query response semantic coding vector will also have the semantic query coding richness of the semantic query coding feature representation based on the feature embedding coding and the feature differentiation representation of the local time domain implicit correlation characteristics of the temperature in the pipe, resulting in repetition of unstructured probability mapping, affecting the accuracy of the decoding results.

[0076] Based on this, when the semantic coding vector of the air flow-temperature query response in the pipe is subjected to decoding probability regression through a decoder-based temperature compensation optimization module, the semantic coding vector of the air flow-temperature query response in the pipe is first optimized, and the optimization process specifically includes:

[0077] Determine the number of zero eigenvalues ​​n in the semantic encoding vector of the air flow-temperature in the duct query response and the zero differentiable value (n-1) of the air flow-temperature in the duct query response obtained by subtracting one from the number of zero eigenvalues ​​n;

[0078] The air flow-in-pipe temperature query response zero differentiable value (n-1) is multiplied and divided by the number of zero eigenvalues ​​n to obtain a first air flow-in-pipe temperature query response field fitting value (n-1)×n and a second air flow-in-pipe temperature query response field fitting value (n-1) / n;

[0079] Calculate the square root of the sum of squares of all eigenvalues ​​of the air flow-duct temperature query response semantic encoding vector to obtain the air flow-duct temperature query response mode characterization value Among them, v i represents each eigenvalue of the semantic encoding vector of the air flow-temperature in the duct query response, Indicates the air flow-duct temperature query response mode characterization value;

[0080] Each eigenvalue v of the semantic encoding vector of the air flow-pipeline temperature query response is calculated using the zero differentiable value (n-1) of the air flow-pipeline temperature query response as an index. i The power function v i n-1 , and multiply it by the first air flow-pipe temperature query response field fitting value (n-1)×n to obtain the first air flow-pipe temperature query response intermediate representation value v 1i ;

[0081] Calculate each eigenvalue v of the semantic encoding vector of the air flow-temperature in the duct query response i The second air flow-pipe temperature query response field fitting value (n-1) / n and the air flow-pipe temperature query response mode characterization value The product of the second air flow rate and the temperature in the duct is obtained by querying the intermediate representation value v of the response 2i ;

[0082] The first air flow-pipe temperature query response intermediate representation value v 1i The second air flow-duct temperature query response intermediate representation value v 2i The weighted difference w1v 1i -w2v 12 The optimized air flow-duct temperature query response semantic encoding vector is composed, where w1 and w2 represent weighted hyperparameters.

[0083] That is, based on the vector field fitting law of high-dimensional manifold, the isolated zeros in the semantic coding vector of the air flow-temperature query in-duct are used as the differentiable dimension of the high-dimensional manifold, so as to fix the normal direction pointing to the distribution center near the local coordinates represented by the eigenvalues ​​of the semantic coding vector of the air flow-temperature query in-duct, thereby avoiding invalid repetition in the probability mapping process caused by the unstructured characteristics of the interactively rich distribution of the semantic coding vector of the air flow-temperature query in-duct due to its effective alignment with the probability density differentiable field, and improving the accuracy of the corrected air flow value obtained by the decoder-based temperature compensation optimization module.

[0084] In summary, the integrated intelligent air flow meter according to the embodiment of the present application is explained, which performs air detection based on the ultrasonic principle, and deploys a pair of ultrasonic transducers in the sewage deodorization pipeline, and performs a preliminary calculation of the air flow based on the influence of air flow on the ultrasonic propagation velocity and in combination with the pipeline parameters. Then, in order to correct the influence of temperature changes on the ultrasonic propagation velocity, the integrated intelligent air flow meter further monitors and analyzes the temperature in the pipeline in real time, and compensates and corrects the preliminarily calculated air flow based on the time series change pattern of the temperature to achieve more accurate air flow measurement. In this way, the measurement error caused by temperature fluctuations can be effectively reduced, and the measurement accuracy and stability of the air flow can be improved.

[0085] The basic principle of the present invention is described above in conjunction with specific embodiments. However, it should be pointed out that the advantages, strengths, effects, etc. mentioned in the present invention are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. must be possessed by each embodiment of the present invention. In addition, the specific details of the above embodiments are only for the purpose of illustration and facilitation of understanding, rather than limitation, and the above details do not limit the present invention to being implemented by adopting the above specific details.

Claims

1. An integrated intelligent air flow meter, characterized in that: include: a first ultrasonic transducer and a second ultrasonic transducer, the first ultrasonic transducer being used to transmit a first ultrasonic detection signal to the second ultrasonic transducer, the second ultrasonic transducer being used to transmit a second ultrasonic detection signal to the first ultrasonic transducer, the first ultrasonic transducer being further used to receive the second ultrasonic detection signal and the second ultrasonic transducer being further used to receive the first ultrasonic detection signal; A temperature acquisition unit, used to acquire the time queue of the temperature in the pipeline through a temperature sensor; A signal processing unit, configured to determine an initial air flow value based on a first time from the transmission to the reception of the first ultrasonic detection signal and a second time from the transmission to the reception of the second ultrasonic detection signal; A signal transmission unit, used for transmitting the time queue of the temperature in the pipeline to the signal processing unit; The signal processing unit is further used to perform temperature compensation on the initial air flow value based on the time sequence of the temperature in the pipeline to obtain a corrected air flow value.

2. The integrated intelligent air flow meter according to claim 1, characterized in that: The signal processing unit comprises: A time difference calculation subunit, configured to calculate the time difference between the first time and the second time to obtain a downstream-upstream time difference; An air average flow rate calculation subunit, used to calculate the air average flow rate value based on the downstream-reverse flow time difference; The initial air flow calculation subunit is used to calculate the product of the pipeline cross-sectional area and the average air flow velocity value to obtain the initial air flow value.

3. The integrated intelligent air flow meter according to claim 2, characterized in that: The air average flow rate calculation subunit is used to: Based on the downstream-reverse flow time difference, the average air flow rate value is calculated using the following flow rate calculation formula, wherein the flow rate calculation formula is: Wherein, L is the diameter of the pipe, d is the distance between the first ultrasonic transducer and the second ultrasonic transducer, θ is the angle between the ultrasonic path and the air flow direction, Δt is the downstream-reverse flow time difference, and v is the average air flow velocity value.

4. The integrated intelligent air flow meter according to claim 3, characterized in that: The signal processing unit comprises: A phase space reconstruction subunit, used for performing phase space reconstruction on the time queue of the temperature in the pipeline to obtain a set of time series subsequences of the temperature in the pipeline; A time series feature extraction subunit, used to extract the time series features of each pipe temperature time series subsequence in the set of pipe temperature time series subsequences to obtain a sequence of implicit coding feature vectors of the local time series of the pipe temperature; An embedded coding subunit, used for embedding the initial air flow value to obtain an initial air flow embedded coding vector; A fast semantic query subunit, used for performing fast scanning semantic query coding on the sequence of the local temporal implicit coding feature vector of the temperature in the pipeline and the initial air flow embedded coding vector to obtain an air flow-temperature in the pipeline query response semantic coding vector; The air flow correction subunit is used to generate the corrected air flow value based on the air flow-in-duct temperature query response semantic coding vector.

5. The integrated intelligent air flow meter according to claim 4, characterized in that: The temporal feature extraction subunit is used to: Each of the pipeline temperature time series subsequences in the set of pipeline temperature time series subsequences is input into a sequence encoder based on the RNN model to obtain a sequence of implicitly encoded feature vectors of the local time series of the pipeline temperature.

6. The integrated intelligent air flow meter according to claim 5, characterized in that: The fast semantic query subunit includes: A fast matching secondary subunit is used to select a subsequence of the local temporal implicit coding feature vectors of the in-pipe temperature from the sequence of the local temporal implicit coding feature vectors of the in-pipe temperature based on the semantic correlation between the initial air flow embedded coding vector and each local temporal implicit coding feature vector of the in-pipe temperature in the sequence of the local temporal implicit coding feature vectors of the in-pipe temperature; The cross-domain query response analysis secondary subunit is used to perform cross-domain query response analysis on the subsequence of the initial air flow embedded coding vector and the local temporal implicit coding feature vector of the temperature in the pipeline to obtain the air flow-pipeline temperature query response semantic coding vector.

7. The integrated intelligent air flow meter according to claim 6, characterized in that: The fast matching secondary subunit is used for: Calculating the mutual information between the initial air flow embedded coding vector and each local temporal implicit coding feature vector of the in-pipe temperature in the sequence of the local temporal implicit coding feature vector of the in-pipe temperature to obtain a sequence of air flow-in-pipe temperature fast semantic matching factors; identifying a first maximum value and a second maximum value from the sequence of air flow-duct temperature fast semantic matching factors; Based on the positions of the first maximum value and the second maximum value in the sequence of the air flow-temperature in the duct fast semantic matching factors, a subsequence of the local temporal implicit coding feature vector of the temperature in the duct that is quickly matched is determined from the sequence of the local temporal implicit coding feature vector of the temperature in the duct.

8. The integrated intelligent air flow meter according to claim 7, characterized in that: The cross-domain query response analysis secondary subunit is used to: The initial air flow embedded coding vector is linearly transformed to obtain a query vector and a value vector, and a subsequence of the local temporal implicit coding feature vector of the temperature in the pipeline is used as a subsequence of the key vector. The query vector, the value vector and the subsequence of the key vector are input into a cross-domain query coding module based on a converter structure to obtain the air flow-pipeline temperature query response semantic coding vector.

9. The integrated intelligent air flow meter according to claim 8, characterized in that: The air flow correction subunit is used for: The air flow-in-duct temperature query response semantic encoding vector is input into a decoder-based temperature compensation optimization module to obtain the corrected air flow value.

Images