CN112504428A - Low-power-consumption embedded high-speed distributed optical fiber vibration sensing system and application thereof - Google Patents
Low-power-consumption embedded high-speed distributed optical fiber vibration sensing system and application thereof Download PDFInfo
- Publication number
- CN112504428A CN112504428A CN202011116317.3A CN202011116317A CN112504428A CN 112504428 A CN112504428 A CN 112504428A CN 202011116317 A CN202011116317 A CN 202011116317A CN 112504428 A CN112504428 A CN 112504428A
- Authority
- CN
- China
- Prior art keywords
- data
- module
- control module
- memory
- optical fiber
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000013307 optical fiber Substances 0.000 title claims abstract description 36
- 238000012545 processing Methods 0.000 claims abstract description 73
- 238000006243 chemical reaction Methods 0.000 claims abstract description 44
- 238000001514 detection method Methods 0.000 claims abstract description 44
- 238000007781 pre-processing Methods 0.000 claims abstract description 33
- 238000000605 extraction Methods 0.000 claims abstract description 29
- 238000012549 training Methods 0.000 claims abstract description 19
- 238000001914 filtration Methods 0.000 claims abstract description 13
- 238000000034 method Methods 0.000 claims abstract description 9
- 230000015654 memory Effects 0.000 claims description 55
- 238000004891 communication Methods 0.000 claims description 30
- 238000001228 spectrum Methods 0.000 claims description 17
- 238000005070 sampling Methods 0.000 claims description 15
- 230000006870 function Effects 0.000 claims description 13
- 230000009545 invasion Effects 0.000 claims description 11
- 238000004458 analytical method Methods 0.000 claims description 9
- 238000013527 convolutional neural network Methods 0.000 claims description 8
- 230000009466 transformation Effects 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000009432 framing Methods 0.000 claims description 6
- 230000003287 optical effect Effects 0.000 claims description 4
- 238000013075 data extraction Methods 0.000 claims description 3
- 238000013461 design Methods 0.000 claims description 3
- 239000000284 extract Substances 0.000 claims description 3
- 238000002372 labelling Methods 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 238000004148 unit process Methods 0.000 claims description 2
- 238000012544 monitoring process Methods 0.000 description 10
- 230000007613 environmental effect Effects 0.000 description 3
- 230000006399 behavior Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 description 2
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000005055 memory storage Effects 0.000 description 1
- 239000003345 natural gas Substances 0.000 description 1
- 239000003208 petroleum Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
- G01H9/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves by using radiation-sensitive means, e.g. optical means
- G01H9/004—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves by using radiation-sensitive means, e.g. optical means using fibre optic sensors
Abstract
The invention relates to a low-power-consumption embedded high-speed distributed optical fiber vibration sensing system and application thereof, which are characterized in that: the upper platform is used for realizing system initialization, intrusion vibration type model training and prediction and alarm output; the FPGA processing unit is used for realizing the control of the data acquisition process and the preprocessing of the acquired data; the DSP unit is used for realizing data characteristic value extraction, intrusion vibration type prediction and alarm output; the detection conversion unit is used for acquiring a laser vibration signal and completing photoelectric conversion; compared with the prior art, the off-chip memory chip is controlled by adopting a ping-pong control mode, so that the data is stored and processed in real time, the data is stored according to a time sequence during storage, and the data is read according to a distance sequence during reading; the FPGA processing unit internally preprocesses the original signals one by one from the distance point location in a pipeline mode, and judges whether the point is disturbed or not in a FIR filtering and threshold value judging mode.
Description
The technical field is as follows:
the invention relates to the technical field of distributed optical fiber vibration sensing, in particular to a low-power-consumption embedded high-speed distributed optical fiber vibration sensing system with strong environmental adaptability, high computing capability and accurate identification of an invasion vibration signal and application thereof.
Background art:
in recent years, security and protection problems become global concerns, and the distributed optical fiber vibration intrusion monitoring system can sense external vibration information, resist electromagnetic interference and influence from severe environment as a novel perimeter intrusion monitoring system, and realize intrusion monitoring of pipelines, stations and important buildings. The method has obvious advantages in distance measurement, service life, safety, reliability, concealment, detection precision and alarm efficiency. At present, monitoring application based on a distributed optical fiber vibration sensing technology is applied to many fields, such as long-distance natural gas, petroleum pipeline monitoring, optical cable monitoring and other safety monitoring; there are also security monitoring of sensitive facilities such as airports, nuclear power plants, factories, and military bases.
However, in the application of such DVS systems, especially in the field of long-distance monitoring, various problems are often encountered, such as: due to the large amount of collected and calculated data, the real-time performance of alarming is difficult to realize; the monitoring environment is complex and changeable, and the invasion behavior of the pipeline is difficult to calculate and judge; the traditional DVS system based on the PC is required to be installed in a cabinet of a machine room, the environmental adaptability is poor, and the installation place is limited, so that the equipment cannot run in the industrial application environment with severe environment, such as high temperature, low temperature, no commercial power supply and the like.
The invention content is as follows:
aiming at the defects and shortcomings in the prior art, the invention provides a low-power-consumption embedded high-speed distributed optical fiber vibration sensing system with strong real-time performance, good environmental adaptability and accurate prediction and identification and application thereof.
The invention is achieved by the following measures:
a low-power-consumption embedded high-speed distributed optical fiber vibration sensing system is characterized by being provided with:
the upper platform is used for realizing system initialization, intrusion vibration type model training and prediction and alarm output;
the FPGA processing unit is used for realizing the control of the data acquisition process and the preprocessing of the acquired data;
the DSP unit is used for realizing data characteristic value extraction, intrusion vibration type prediction and alarm output;
the detection conversion unit is used for acquiring a laser vibration signal and completing photoelectric conversion;
the upper platform is connected with the DSP unit through the Ethernet, the DSP unit and the FPGA processing unit complete data communication through an SRIO protocol, and the output end of the detection conversion unit is connected with the FPGA processing unit;
the detection conversion unit is internally provided with a light path detection module, a photoelectric conversion module and an analog-to-digital conversion module which are sequentially connected, wherein the output end of the analog-to-digital conversion module is connected with the FPGA processing unit; the FPGA processing unit is provided with a detection acquisition control module, an SRIO communication module, a data preprocessing module, a ping-pong control module and two off-chip memories connected with the ping-pong control module, wherein the control signal output end of the detection acquisition control module is connected with the light path detection module of the detection conversion unit, the SRIO communication module is respectively connected with the detection acquisition control module, the data preprocessing module and the DSP unit, the input end of the ping-pong control module receives data output by the analog-digital conversion module in the detection conversion unit and stores the data into the two off-chip memories, and the data preprocessing module extracts the data in the two off-chip memories through the ping-pong control module for preprocessing and then stores the data into the DSP unit through the SRIO module;
the data preprocessing module is sequentially provided with an FIR filtering processing module, a data square summation module, an energy value and threshold value comparison module and a position point labeling processing module;
the DSP unit is internally provided with an SRIO communication module, a characteristic value extraction module, an invasion vibration type prediction module, an alarm output module and an Ethernet communication module, wherein the characteristic value extraction module is respectively connected with the Ethernet communication module and the invasion vibration type prediction module, the output end of the invasion vibration type prediction module is connected with the alarm output module, and the characteristic value extraction module is internally provided with a data position point and original data extraction module, a pre-emphasis and framing windowing module, a fast Fourier transform processing module, a Mel filter, a cepstrum analysis module and a Mel frequency cepstrum coefficient output module which are sequentially connected; and a convolutional neural network model is arranged in the intrusion vibration type prediction module.
The invention also provides application of the low-power-consumption embedded high-speed distributed optical fiber vibration sensing system, which is characterized by comprising the following steps of:
step 1: initializing, wherein a detection conversion module acquires an optical fiber vibration signal through a light path detection module, converts the optical fiber vibration signal into an electric signal and uploads the electric signal to an FPGA (field programmable gate array) processing unit;
step 2: the FPGA processing unit utilizes the ping-pong control module to store and preprocess the data uploaded by the detection conversion module, and uploads the preprocessed data to the DSP unit, and the preprocessing comprises the following steps:
step 2-1: performing FIR filtering processing on the data;
step 2-2: performing square summation on the data after FIR filtering processing;
step 2-3: comparing the data energy value obtained by the square summation with a threshold value, and if the data energy value is smaller than the threshold value, not processing; if the distance is greater than the threshold value, marking a position point before the data, wherein the position point is used for recording the spatial distance information, and temporarily storing the data to wait for sending the data to the DSP unit;
and step 3: and the DSP unit is used for carrying out signal feature extraction, intrusion vibration type prediction and alarm output on the preprocessed data.
The initialization in step 1 of the present invention includes: the upper platform, the FPGA processing unit and the DSP unit load respective programs, complete SRIO communication and Ethernet communication initialization, establish communication among the programs, and then the upper platform firstly configures parameters of the system through the Ethernet, wherein the parameters comprise laser pulse frequency, pulse width, sampling frequency, acquisition length and threshold value setting, are stored in the DSP unit and are simultaneously issued to the FPGA processing unit; the FPGA processing unit controls the detection acquisition module to acquire data, and converts optical signals into digital signals through the photoelectric conversion module and the analog-to-digital conversion module.
When the FPGA processing unit stores data in the off-chip memory in the step 2, the two memories adopt a ping-pong design and are controlled by a ping-pong control module, the ping-pong control module comprises a data control module 1 and a data control module 2, the two modules work alternately and perform interlocking processing, the data control module 1 controls to write data into the memory 1 and read data from the memory 2; the data control module 2 controls the data writing into the memory 2, reads the data from the memory 1, and stores and reads the data with fixed time length each time; when the collected data are stored, the ping-pong control module controls the data to be stored in the memory, the data of single laser pulse are stored according to the sequence of spatial distance, and the laser pulses are stored according to the time sequence to form a matrix containing time information and spatial information;
when the FPGA processing unit processes the acquired data in the step 2, the ping-pong control module reads the data with fixed time length of the same position point in the memory according to the spatial distance sequence and transmits the data into the data preprocessing module; the method comprises the steps that the working state of a data control module is judged according to whether a data selection bit parameter is 0 or not, for example, the data selection bit is 0, the data control module 1 works, an FPGA sends out a write enable signal of a memory 1 and a read enable signal of a memory 2, and then a plurality of parameters, namely a single sampling point number, a position point and an m, are set, so that the data of the same position point on a time dimension can be read when the data of the memory 2 are read, and the data can be processed; the single sampling point number is a discrete sampling point number corresponding to the laser pulse, the position point is a point number of a distance position of the original data corresponding to the laser pulse in space, m is a count, whether the stored and read data exceed the upper limit of the memory or not is recorded, the acquired data are stored in the memory 1 according to a time sequence, so that the address is +1 in sequence, the data in the same position point within a period of time are read by the read data, the single sampling point number is required to be sequentially added to the address of the memory, and before the position point exceeds the upper limit of the storage space of the memory, the address is reset to be the next position point, and the data of the next position point are continuously read.
Data in a memory controlled by the DSP unit in the step 3 are transmitted to an upper platform through an Ethernet module, and an invasive vibration type model is obtained through characteristic value extraction, invasive vibration type model training and prediction module training, wherein the characteristic value extraction method is Mel frequency spectrum transformation, a Mel frequency cepstrum coefficient is extracted to serve as a characteristic value, and the invasive vibration type model training is carried out by using a convolutional neural network model to obtain an invasive vibration type prediction model; the Mel frequency spectrum transformation calculation process specifically comprises the following steps:
firstly, extracting a position point at the beginning of original data, then carrying out pre-emphasis, framing and windowing on the original data, dividing a signal in a time period into a plurality of short time analysis windows, respectively carrying out fast Fourier transform, converting the obtained frequency spectrum into a Mel frequency spectrum through a Mel filter, wherein the calculation formula is as follows: mel (f) ═ 2595 log10(1+ f700), performing cepstrum analysis on the Mel frequency spectrum by using DCT discrete cosine transform to obtain Mel frequency cepstrum coefficient MFCC as a characteristic value, and performing next model training and model prediction.
The invention also comprises a step 4: the intrusion vibration type prediction model is issued to the DSP through the Ethernet module for storage; the subsequent working state can be divided into two types, one is an operation mode of an upper platform when the upper platform is separated, and the functions of data acquisition, data preprocessing, characteristic value extraction, intrusion vibration type prediction, alarm output and the like are independently completed by a low-power-consumption embedded high-speed distributed optical fiber vibration sensing system; the functions of data acquisition and data preprocessing and the like are completed by a low-power-consumption embedded high-speed distributed optical fiber vibration sensing system by means of the operation mode of an upper platform, and the functions of characteristic value extraction, intrusion vibration type prediction and alarm output are completed by the upper platform.
Compared with the prior art, the invention has the following advantages: (1) an embedded structure is adopted to realize a distributed optical fiber vibration sensing system; (2) the off-chip memory chip is controlled by adopting a ping-pong control mode, so that real-time data storage and processing are realized, the data are stored according to a time sequence during storage, and the data are read according to a distance sequence during reading; (3) the FPGA processing unit internally preprocesses the original signals one by one from the distance point location in a pipeline mode, judges whether the point is disturbed or not in a FIR filtering and threshold value judging mode, and temporarily stores the disturbed point in FIFO to the point without disturbance. (4) Using SRIO protocol to complete high-speed chain DMA data transmission between FPGA and DSP; (5) extracting a characteristic value of a disturbance point by using a Mel frequency spectrum, and performing model training and prediction on the characteristic value by using a convolutional neural network; (6) after parameter setting and model downloading are completed, the low-power-consumption embedded high-speed distributed optical fiber vibration sensing system can independently complete data acquisition, processing and alarming functions, and alarming information can be transmitted to other platforms through Ethernet and 485 communication; and the upper platform can also realize the functions of data processing and alarming.
Description of the drawings:
FIG. 1 is a block diagram of the present invention.
FIG. 2 is a flow chart of the FPGA processing unit storing data according to the present invention.
FIG. 3 is a flow chart of the FPGA processing unit of the present invention for pre-processing data.
FIG. 4 is a flow chart of the eigenvalue extraction process in the DSP unit of the present invention.
The specific implementation mode is as follows:
the invention is further described below with reference to the accompanying drawings and examples.
As shown in fig. 1, the present invention provides a low power consumption embedded high-speed distributed optical fiber vibration sensing system, which is provided with: the upper platform is used for realizing system initialization, intrusion vibration type model training and prediction and alarm output;
the FPGA processing unit is used for realizing the control of the data acquisition process and the preprocessing of the acquired data;
the DSP unit is used for realizing data characteristic value extraction, intrusion vibration type prediction and alarm output;
the detection conversion unit is used for acquiring a laser vibration signal and completing photoelectric conversion;
the upper platform is connected with the DSP unit through the Ethernet, the DSP unit and the FPGA processing unit complete data communication through an SRIO protocol, and the output end of the detection conversion unit is connected with the FPGA processing unit;
the detection conversion unit is internally provided with a light path detection module, a photoelectric conversion module and an analog-to-digital conversion module which are sequentially connected, wherein the output end of the analog-to-digital conversion module is connected with the FPGA processing unit; the FPGA processing unit is provided with a detection acquisition control module, an SRIO communication module, a data preprocessing module, a ping-pong control module and two off-chip memories connected with the ping-pong control module, wherein the control signal output end of the detection acquisition control module is connected with the light path detection module of the detection conversion unit, the SRIO communication module is respectively connected with the detection acquisition control module, the data preprocessing module and the DSP unit, the input end of the ping-pong control module receives data output by the analog-digital conversion module in the detection conversion unit and stores the data into the two off-chip memories, and the data preprocessing module extracts the data in the two off-chip memories through the ping-pong control module for preprocessing and then stores the data into the DSP unit through the SRIO module;
the data preprocessing module is sequentially provided with an FIR filtering processing module, a data square summation module, an energy value and threshold value comparison module and a position point labeling processing module;
the DSP unit is internally provided with an SRIO communication module, a characteristic value extraction module, an invasion vibration type prediction module, an alarm output module and an Ethernet communication module, wherein the characteristic value extraction module is respectively connected with the Ethernet communication module and the invasion vibration type prediction module, the output end of the invasion vibration type prediction module is connected with the alarm output module, and the characteristic value extraction module is internally provided with a data position point and original data extraction module, a pre-emphasis and framing windowing module, a fast Fourier transform processing module, a Mel filter, a cepstrum analysis module and a Mel frequency cepstrum coefficient output module which are sequentially connected; and a convolutional neural network model is arranged in the intrusion vibration type prediction module.
The invention also provides application of the low-power-consumption embedded high-speed distributed optical fiber vibration sensing system, which is characterized by comprising the following steps of:
step 1: initializing, wherein a detection conversion module acquires an optical fiber vibration signal through a light path detection module, converts the optical fiber vibration signal into an electric signal and uploads the electric signal to an FPGA (field programmable gate array) processing unit;
step 2: the FPGA processing unit utilizes the ping-pong control module to store and preprocess the data uploaded by the detection conversion module, and uploads the preprocessed data to the DSP unit, and the preprocessing comprises the following steps:
step 2-1: performing FIR filtering processing on the data;
step 2-2: performing square summation on the data after FIR filtering processing;
step 2-3: comparing the data energy value obtained by the square summation with a threshold value, and if the data energy value is smaller than the threshold value, not processing; if the distance is larger than the threshold value, adding a position point before the data, wherein the position point is used for recording the spatial distance information, and temporarily storing the data to wait for sending the data to the DSP unit;
and step 3: and the DSP unit is used for carrying out signal feature extraction, intrusion vibration type prediction and alarm output on the preprocessed data.
Example 1:
the embodiment provides an application of a low-power-consumption embedded high-speed distributed optical fiber vibration sensing system, which specifically comprises the following contents:
step 1: when the system is powered on for the first time, the FPGA processing unit and the DSP unit load respective programs, complete SRIO communication and Ethernet communication initialization, and establish communication between the two units;
step 2: the upper platform firstly carries out parameter configuration on the system through the Ethernet, wherein the parameter configuration comprises laser pulse frequency, pulse width, sampling frequency, acquisition length, threshold value setting and the like, the parameters are stored in a DSP chip and are simultaneously issued to an FPGA processing unit;
and step 3: the FPGA processing unit controls the detection acquisition module to acquire data, and converts an optical signal into a digital signal through the photoelectric conversion module and the analog-to-digital conversion module;
because the data volume is large and real-time processing is required, the two memories adopt a ping-pong design and are controlled by a ping-pong control module; the ping-pong control module comprises a data control module 1 and a data control module 2 which work alternately and perform interlocking treatment. The data control module 1 controls the data writing into the memory 1 and the data reading from the memory 2; the data control module 2 controls the data writing into the memory 2 and the data reading from the memory 1; storing and reading data with fixed time length each time; when the collected data are stored, the ping-pong control module controls the data to be stored in the memory, the data of single laser pulse are stored according to the sequence of spatial distance, and the laser pulses are stored according to the time sequence to form a matrix containing time information and spatial information;
and 4, step 4: when the collected data are processed, the ping-pong control module reads the data with fixed time length of the same position point in the memory according to the spatial distance sequence and transmits the data into the data preprocessing module; as shown in fig. 2 and fig. 3 of the flowcharts for collecting, storing and reading original data, firstly, the working state of the data control module is determined according to whether the parameter of the data selection bit is 0, for example, the data selection bit is 0, the data control module 1 works, the FPGA processing unit sends out a write enable signal of the memory 1 and a read enable signal of the memory 2, and then sets parameters, a single sampling point number, a position point, and an m, which are all used to ensure that data of the same position point in a time dimension can be read when the data of the memory 2 is read, so as to perform the next data processing. The number of the single sampling points is the number of discrete sampling points corresponding to the laser pulse, the position point is the number of points of the original data corresponding to the distance position of the laser pulse in the space, m is a count, whether the stored and read data exceed the upper limit of the memory or not is recorded, and the acquired data are stored in the memory 1 according to the time sequence, so that the addresses are +1 in sequence; reading data in the same position point for a period of time requires that the memory address is sequentially increased by the number of single sampling points, and before the position point exceeds the upper limit of the memory storage space, the address is reset to the next position point, and the data of the next position point is continuously read.
As shown in fig. 3, FIR filtering is performed on the raw data of each spatial position point read by the ping-pong control module in a pipeline manner in the data preprocessing module, each spatial position point obtains a set of one-dimensional number sequences, each element is squared and accumulated, and the obtained value is the signal energy value of the spatial position point and is compared with the threshold value in the parameter setting. If the distance is greater than the threshold value, it is indicated that the distance position point corresponding to the group of data has an intrusion disturbance behavior, the group of data columns are temporarily stored in an FIFO storage space inside the FPGA, and 16-bit data recording space distance information is added at the beginning of the data columns; if the threshold value is less than the threshold value, the next processing is not carried out. All the distance points are calculated according to the steps, and the FPGA can perform parallel data processing, so that a pipeline processing mode is designed;
after a group of data with fixed time length is processed, transmitting the data temporarily stored in the FIFO storage space in the FPGA to a DSP by a chain DMA through an SRIO communication module and storing the data in a memory chip;
and 5: data in a memory controlled by a DSP unit is transmitted to an upper computer through an Ethernet module, an invasive vibration type model is obtained through characteristic value extraction, invasive vibration type model training and prediction module training, wherein the characteristic value extraction method is Mel frequency spectrum transformation, Mel frequency cepstrum coefficients are extracted to serve as characteristic values, the invasive vibration type model training uses convolutional neural network model training to obtain an invasive vibration type prediction model, the Mel frequency spectrum transformation calculation process is shown as a flow chart 4, firstly, position points at the beginning of original data are extracted, then, pre-emphasis, framing and windowing are carried out on the original data, signals in a time period are divided into a plurality of short analysis windows, fast Fourier transformation is respectively carried out, obtained frequency spectrums are converted into Mel frequency spectrums through a Mel filter, and the calculation formula is as follows:
MEL(f)＝2595*log10(1+ f700), performing cepstrum analysis on the Mel frequency spectrum by using DCT discrete cosine transform to obtain a Mel frequency cepstrum coefficient MFCC as a characteristic value, and performing next model training and model prediction;
step 6: the intrusion vibration type prediction model is issued to the DSP through the Ethernet module for storage; and 7: the subsequent working state can be divided into two types, one is an operation mode of an upper platform when the upper platform is separated, and the functions of data acquisition, data preprocessing, characteristic value extraction, intrusion vibration type prediction, alarm output and the like are independently completed by a low-power-consumption embedded high-speed distributed optical fiber vibration sensing system; the functions of data acquisition and data preprocessing and the like are completed by a low-power-consumption embedded high-speed distributed optical fiber vibration sensing system by means of the operation mode of an upper platform, and the functions of characteristic value extraction, intrusion vibration type prediction and alarm output are completed by the upper platform.
Compared with the prior art, the invention has the following advantages: (1) an embedded structure is adopted to realize a distributed optical fiber vibration sensing system; (2) the off-chip memory chip is controlled by adopting a ping-pong control mode, so that real-time data storage and processing are realized, the data are stored according to a time sequence during storage, and the data are read according to a distance sequence during reading; (3) the FPGA processing unit internally preprocesses the original signals one by one from the distance point location in a pipeline mode, judges whether the point is disturbed or not in a FIR filtering and threshold value judging mode, and temporarily stores the disturbed point in FIFO to the point without disturbance. (4) Using SRIO protocol to complete high-speed chain DMA data transmission between FPGA and DSP; (5) extracting a characteristic value of a disturbance point by using a Mel frequency spectrum, and performing model training and prediction on the characteristic value by using a convolutional neural network; (6) after parameter setting and model downloading are completed, the low-power-consumption embedded high-speed distributed optical fiber vibration sensing system can independently complete data acquisition, processing and alarming functions, and alarming information can be transmitted to other platforms through Ethernet and 485 communication; and the upper platform can also realize the functions of data processing and alarming.
Claims (7)
1. A low-power-consumption embedded high-speed distributed optical fiber vibration sensing system is characterized by being provided with:
the upper platform is used for realizing system initialization, intrusion vibration type model training and prediction and alarm output;
the FPGA processing unit is used for realizing the control of the data acquisition process and the preprocessing of the acquired data;
the DSP unit is used for realizing data characteristic value extraction, intrusion vibration type prediction and alarm output;
the detection conversion unit is used for acquiring a laser vibration signal and completing photoelectric conversion;
the upper platform is connected with the DSP unit through the Ethernet, the DSP unit and the FPGA processing unit complete data communication through an SRIO protocol, and the output end of the detection conversion unit is connected with the FPGA processing unit;
the detection conversion unit is internally provided with a light path detection module, a photoelectric conversion module and an analog-to-digital conversion module which are sequentially connected, wherein the output end of the analog-to-digital conversion module is connected with the FPGA processing unit; the FPGA processing unit is provided with a detection acquisition control module, an SRIO communication module, a data preprocessing module, a ping-pong control module and two off-chip memories connected with the ping-pong control module, wherein the control signal output end of the detection acquisition control module is connected with the light path detection module of the detection conversion unit, the SRIO communication module is respectively connected with the detection acquisition control module, the data preprocessing module and the DSP unit, the input end of the ping-pong control module receives data output by the analog-digital conversion module in the detection conversion unit and stores the data into the two off-chip memories, and the data preprocessing module extracts the data in the two off-chip memories through the ping-pong control module for preprocessing and then stores the data into the DSP unit through the SRIO module;
the data preprocessing module is sequentially provided with an FIR filtering processing module, a data square summation module, an energy value and threshold value comparison module and a position point labeling processing module;
the DSP unit is internally provided with an SRIO communication module, a characteristic value extraction module, an invasion vibration type prediction module, an alarm output module and an Ethernet communication module, wherein the characteristic value extraction module is respectively connected with the Ethernet communication module and the invasion vibration type prediction module, the output end of the invasion vibration type prediction module is connected with the alarm output module, and the characteristic value extraction module is internally provided with a data position point and original data extraction module, a pre-emphasis and framing windowing module, a fast Fourier transform processing module, a Mel filter, a cepstrum analysis module and a Mel frequency cepstrum coefficient output module which are sequentially connected; and a convolutional neural network model is arranged in the intrusion vibration type prediction module.
2. The application of the low-power embedded high-speed distributed optical fiber vibration sensing system according to claim 1, which comprises the following steps:
step 1: initializing, wherein a detection conversion module acquires an optical fiber vibration signal through a light path detection module, converts the optical fiber vibration signal into an electric signal and uploads the electric signal to an FPGA (field programmable gate array) processing unit;
step 2: the FPGA processing unit utilizes the ping-pong control module to store and preprocess the data uploaded by the detection conversion module, and uploads the preprocessed data to the DSP unit, and the preprocessing comprises the following steps:
step 2-1: performing FIR filtering processing on the data;
step 2-2: performing square summation on the data after FIR filtering processing;
step 2-3: comparing the data energy value obtained by the square summation with a threshold value, and if the data energy value is smaller than the threshold value, not processing; if the distance is larger than the threshold value, adding a position point before the data, wherein the position point is used for recording the spatial distance information, and temporarily storing the data to wait for sending the data to the DSP unit;
and step 3: and the DSP unit is used for carrying out signal feature extraction, intrusion vibration type prediction and alarm output on the preprocessed data.
3. The application of the low-power-consumption embedded high-speed distributed optical fiber vibration sensing system according to claim 2, wherein the initialization in step 1 comprises: the upper platform, the FPGA processing unit and the DSP unit load respective programs, complete SRIO communication and Ethernet communication initialization, establish communication among the programs, and then the upper platform firstly configures parameters of the system through the Ethernet, wherein the parameters comprise laser pulse frequency, pulse width, sampling frequency, acquisition length and threshold value setting, are stored in the DSP unit and are simultaneously issued to the FPGA processing unit; the FPGA processing unit controls the detection acquisition module to acquire data, and converts optical signals into digital signals through the photoelectric conversion module and the analog-to-digital conversion module.
4. The application of the low-power-consumption embedded high-speed distributed optical fiber vibration sensing system according to claim 2, wherein when the FPGA processing unit stores data in the off-chip memory in step 2, the two memories adopt a ping-pong design and are controlled by a ping-pong control module because of large data volume and real-time processing requirement, the ping-pong control module comprises a data control module 1 and a data control module 2, the two modules work alternately and perform interlocking processing, the data control module 1 controls to write data into the memory 1 and read data from the memory 2; the data control module 2 controls the data writing into the memory 2, reads the data from the memory 1, and stores and reads the data with fixed time length each time; when the collected data are stored, the ping-pong control module controls the data to be stored in the memory, the data of the single laser pulse are stored according to the sequence of the spatial distance, and the laser pulses are stored according to the time sequence to form a matrix containing time information and spatial information.
5. The application of the low-power-consumption embedded high-speed distributed optical fiber vibration sensing system according to claim 2, wherein when the FPGA processing unit processes the acquired data in step 2, the ping-pong control module reads the data with the set time length of the same position point in the memory according to the spatial distance sequence, and transmits the data into the data preprocessing module; the acquisition, storage and reading of original data firstly judge the working state of a data control module according to whether the parameter of a data selection bit is 0, if the parameter of the data selection bit is 0, the data control module 1 works, an FPGA sends out a write enable signal of a memory 1 and a read enable signal of a memory 2, and then sets parameters, a single sampling point number, a position point and a m, so as to ensure that the data of the same position point on a time dimension can be read when the data of the memory 2 is read, and the data can be processed as the next data; the single sampling point number is a discrete sampling point number corresponding to the laser pulse, the position point is a point number of a distance position of the original data corresponding to the laser pulse in space, m is a count, whether the stored and read data exceed the upper limit of a memory or not is recorded, the acquired data are stored in the memory 1 according to a time sequence, so that the address is +1 in sequence, the data in the same position point within a period of time are read by the read data, the single sampling point number is sequentially increased by the address of the memory at the moment, and before the position point exceeds the upper limit of the storage space of the memory, the address is reset to be the next position point, and the data of the next position point are continuously read.
6. The application of the low-power-consumption embedded high-speed distributed optical fiber vibration sensing system according to claim 2, wherein the data in the memory controlled by the DSP unit in the step 3 is transmitted to an upper platform through an Ethernet module, and an intrusion vibration type model is obtained through characteristic value extraction, intrusion vibration type model training and prediction module training, wherein the characteristic value extraction method is Mel frequency spectrum transformation, Mel frequency cepstrum coefficients are extracted as characteristic values, and the intrusion vibration type model training is carried out by using a convolutional neural network model to obtain an intrusion vibration type prediction model; the Mel frequency spectrum transformation calculation process specifically comprises the following steps:
firstly, extracting a position point at the beginning of original data, then carrying out pre-emphasis, framing and windowing on the original data, dividing a signal in a time period into a plurality of short time analysis windows, respectively carrying out fast Fourier transform, converting the obtained frequency spectrum into a Mel frequency spectrum through a Mel filter, wherein the calculation formula is as follows: mel (f) ═ 2595 log10(1+ f/700), performing cepstrum analysis on the Mel frequency spectrum by using DCT discrete cosine transform to obtain Mel frequency cepstrum coefficient MFCC as a characteristic value, and performing next model training and model prediction.
7. The application of the low-power-consumption embedded high-speed distributed optical fiber vibration sensing system according to claim 2, further comprising the step 4 of: the intrusion vibration type prediction model is issued to the DSP through the Ethernet module for storage; the subsequent working states are divided into two types, one is an operation mode of an upper platform when the upper platform is separated, and the functions of data acquisition, data preprocessing, characteristic value extraction, intrusion vibration type prediction and alarm output are independently completed by a low-power-consumption embedded high-speed distributed optical fiber vibration sensing system; the functions of data acquisition and data preprocessing and the like are completed by a low-power-consumption embedded high-speed distributed optical fiber vibration sensing system by means of the operation mode of an upper platform, and the functions of characteristic value extraction, intrusion vibration type prediction and alarm output are completed by the upper platform.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011116317.3A CN112504428A (en) | 2020-10-19 | 2020-10-19 | Low-power-consumption embedded high-speed distributed optical fiber vibration sensing system and application thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011116317.3A CN112504428A (en) | 2020-10-19 | 2020-10-19 | Low-power-consumption embedded high-speed distributed optical fiber vibration sensing system and application thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112504428A true CN112504428A (en) | 2021-03-16 |
Family
ID=74953958
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011116317.3A Pending CN112504428A (en) | 2020-10-19 | 2020-10-19 | Low-power-consumption embedded high-speed distributed optical fiber vibration sensing system and application thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112504428A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115479219A (en) * | 2022-09-20 | 2022-12-16 | 无锡科晟光子科技有限公司 | Intelligent pipeline state monitoring method and device and intelligent pipeline system |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1014477A2 (en) * | 1998-12-22 | 2000-06-28 | Mitsubishi Denki Kabushiki Kaisha | Directing an antenna to receive digital television signals |
| EP1271506A2 (en) * | 2001-06-28 | 2003-01-02 | STMicroelectronics, Inc. | Circuit and method for detecting a servo wedge on spin up of a data-storage disk |
| CN2530301Y (en) * | 2001-03-09 | 2003-01-08 | 北京卡迪欧医疗设备有限责任公司 | Analyser for electrocardiosignal |
| CN103063290A (en) * | 2012-12-14 | 2013-04-24 | 上海华魏光纤传感技术有限公司 | Novel real-time data acquisition and signal process device used for optical fiber vibration measurement system and implementation method thereof |
| CN105045763A (en) * | 2015-07-14 | 2015-11-11 | 北京航空航天大学 | FPGA (Field Programmable Gata Array) and multi-core DSP (Digital Signal Processor) based PD (Pulse Doppler) radar signal processing system and parallel realization method therefor |
| CN110823356A (en) * | 2019-10-09 | 2020-02-21 | 威海北洋光电信息技术股份公司 | Distributed optical fiber intrusion detection method based on Mel frequency spectrum |
| CN111157099A (en) * | 2020-01-02 | 2020-05-15 | 河海大学常州校区 | Distributed optical fiber sensor vibration signal classification method and identification classification system |
| CN111738439A (en) * | 2020-07-21 | 2020-10-02 | 电子科技大学 | Artificial intelligence processing method and processor supporting online learning |
-
2020
- 2020-10-19 CN CN202011116317.3A patent/CN112504428A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1014477A2 (en) * | 1998-12-22 | 2000-06-28 | Mitsubishi Denki Kabushiki Kaisha | Directing an antenna to receive digital television signals |
| CN2530301Y (en) * | 2001-03-09 | 2003-01-08 | 北京卡迪欧医疗设备有限责任公司 | Analyser for electrocardiosignal |
| EP1271506A2 (en) * | 2001-06-28 | 2003-01-02 | STMicroelectronics, Inc. | Circuit and method for detecting a servo wedge on spin up of a data-storage disk |
| CN103063290A (en) * | 2012-12-14 | 2013-04-24 | 上海华魏光纤传感技术有限公司 | Novel real-time data acquisition and signal process device used for optical fiber vibration measurement system and implementation method thereof |
| CN105045763A (en) * | 2015-07-14 | 2015-11-11 | 北京航空航天大学 | FPGA (Field Programmable Gata Array) and multi-core DSP (Digital Signal Processor) based PD (Pulse Doppler) radar signal processing system and parallel realization method therefor |
| CN110823356A (en) * | 2019-10-09 | 2020-02-21 | 威海北洋光电信息技术股份公司 | Distributed optical fiber intrusion detection method based on Mel frequency spectrum |
| CN111157099A (en) * | 2020-01-02 | 2020-05-15 | 河海大学常州校区 | Distributed optical fiber sensor vibration signal classification method and identification classification system |
| CN111738439A (en) * | 2020-07-21 | 2020-10-02 | 电子科技大学 | Artificial intelligence processing method and processor supporting online learning |
Non-Patent Citations (1)
| Title |
|---|
| 刘德全著: "《空中目标ISAR实时成像技术研究》", 30 November 2014 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115479219A (en) * | 2022-09-20 | 2022-12-16 | 无锡科晟光子科技有限公司 | Intelligent pipeline state monitoring method and device and intelligent pipeline system |
| CN115479219B (en) * | 2022-09-20 | 2024-03-01 | 无锡科晟光子科技有限公司 | Intelligent pipeline state monitoring method, monitoring device and intelligent pipeline system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113283344B (en) | Mining conveyor belt deviation detection method based on semantic segmentation network | |
| CN100483144C (en) | Corona detecting method and system for insulator remote surface on optical amplifier | |
| CN1017568B (en) | Device and method for storage characteristic parameter of locally area network system | |
| CN107705515A (en) | A kind of geological disaster monitoring system based on big data | |
| CN104200620A (en) | Building healthy remote monitoring system and building healthy remote monitoring method | |
| CN109001997B (en) | Anti-interference environment information acquisition system | |
| CN107036659A (en) | A kind of Multifunctional environment monitoring sampling apparatus and its method | |
| CN112504428A (en) | Low-power-consumption embedded high-speed distributed optical fiber vibration sensing system and application thereof | |
| CN107784796A (en) | A kind of Intelligent environment forecasting system based on zigbee | |
| CN111352977A (en) | Time sequence data monitoring method based on self-attention bidirectional long-short term memory network | |
| CN112349057A (en) | Deep learning-based indoor smoke and fire detection method | |
| CN115190191B (en) | Power grid industrial control system and control method based on protocol analysis | |
| CN201166711Y (en) | System for monitoring field unattended seismographic station circumstance | |
| CN109768625A (en) | A kind of electric system overhaul management terminal and method | |
| CN104089652A (en) | On-line monitoring system and method of fiber grating transformer | |
| CN106154938A (en) | A kind of shield-tunneling construction process status data logging plant | |
| CN113031055A (en) | Space steel structure vibration response capturing and monitoring system based on earthquake early warning | |
| CN109581984A (en) | Manned equipment running status monitors system and method | |
| CN114739660A (en) | Gearbox health monitoring method and system | |
| CN205451354U (en) | Optic fibre intelligence anti -theft device | |
| CN203783820U (en) | Vibrational state monitoring device for wind generation set with distributed memory function | |
| Hayes et al. | A real-time stream storage and analysis platform for underwater acoustic monitoring | |
| CN117033983B (en) | Unmanned ship self-noise detection and identification method and system | |
| CN116692393A (en) | Underground belt conveyor state monitoring system and method | |
| CN102035253A (en) | Memorable and learnable test method and device for power grid video and environment monitoring system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210316 |