A multi-point maximum demodulation algorithm based on microwave photon filter sensing
By using multi-point maximum and minimum value demodulation algorithms to process swept frequency data in microwave photon sensing systems, the problem of noise interference is solved, the accuracy of temperature and stress sensing and the fit of the system are improved, and the fit of the system is reduced. Experimental cost.
Patent Information
- Application Number
- CN202111487206.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-11-22
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2041-11-22
AI Technical Summary
The presence of noise sources such as spontaneous radiation and photodiodes in existing microwave photon sensing systems results in the direct acquisition of RF power data containing a large amount of noise, making it difficult to accurately extract temperature and stress to be measured without using expensive vector network analyzers. information.
A multi-point maximum demodulation algorithm and multi-point minimum demodulation algorithm based on microwave photon filter sensing are proposed. By processing the original swept data collected by the electric spectrum analyzer, the information to be measured is extracted. The multi-point maximum and minimum values are used to calculate the RF resonance frequency to form a temperature or stress sensing database.
It effectively reduces noise interference, improves the fit and sensitivity of the system, improves the accuracy of temperature and stress sensing, and reduces experimental costs.
Smart Images

Figure CN114236816B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of microwave photonics, microwave photon filters and the application of microwave photon filters in sensing, in particular to a multi-point maximum value demodulation algorithm based on microwave photon filter sensing. Background Art
[0002] In recent years, due to the rapid development of optoelectronics and optical communication technology, the demand for high-frequency electrical signal generation and high-frequency electrical information processing has gradually increased. A new interdisciplinary subject, microwave photonics (MWP), has emerged. [1] Microwave photonic technology has been widely studied in the fields of optical wireless communication systems, radar navigation, etc. [2 1] With the development of MWP technology, some researchers have found that the demodulation of optical fiber sensing systems can be achieved by using MWP technology. [5] Microwave photon filter (MPF) is a common demodulation method of MWP technology. It changes the traditional optical sensing principle from the conversion of temperature and stress to resonant wavelength to the conversion from the measured value to the microwave resonant frequency. By measuring the change of MPF resonant frequency, the demodulation process of temperature and stress to be measured is completed. The key technology of microwave photon sensing based on MPF principle lies in the test of resonant frequency. According to existing literature, most researchers use vector network analyzer (VNA) for sensing experiments and data analysis. [6 7] However, VNA is expensive, making it difficult to apply in engineering practice. Therefore, it is necessary to use the raw frequency sweep data collected by the electrical spectrum analyzer (ESA) to extract the temperature and stress to be measured. However, there are many noise sources in the MWP sensor system, such as amplified spontaneous emission (ASE) light source and photodiode (PD) [8] , which will make the directly collected RF power data contain a lot of noise. Therefore, it is necessary to combine the characteristics of microwave photon temperature or stress sensing and study the signal processing algorithm for the original swept frequency data without using VNA.
[0003] Therefore, the present invention proposes a multi-point maximum demodulation algorithm based on microwave photon filter sensing to process the original frequency sweep data collected by the electrical spectrum analyzer of the microwave photon filter sensing system.
[0004] References
[0005] [1]Yao J P. Microwave photonics[J]. IEEE Journal of Lightwave Technology, 2009, 27(3): 314-335.
[0006] [2] Xing Yunlu, Li Shangyuan, Xue Xiaoxiao, et al. Research on high-frequency broadband distributed coherent microwave photon imaging radar[J]. Chinese Journal of Lasers, 2021, 48(15): 1517003.
[0007] [3]Yao J P. Microwave Photonics: Arbitrary Waveform Generation[J]. Nature Photonics, 2010, 4(2): 79-80.
[0008] [4] Zou XH, BAI Wenlin, Chen Wen, et al.Microwave Photonics for FeaturedApplications in High-Speed Railways: Communications, Detection, and Sensing[J]. IEEE Journal or Lightwave Technology, 2018, 36(19): 4337-4346.
[0009] [5]Ricchiuti AL, Barrera D, Sales S, et al. Long fiber Bragg gratingsensor interrogation using discrete-time Microwave photonic filteringtechniques[J]. Optics Express, 2013, 21(23): 28175-28181.
[0010] [6]Li SY.Wu R, Wang H, etal.Fiber-Ring-Based RF Filtering for Temperature and Transversal Loading Measurement and Interrogation[J]. IEEESensors Journal, 2018, 8(14): 5794-5798.
[0011] [7] Li YX, Zhou CM, Fan D, et al. Iteration Bayesian Reweighed Algorithm for Optical Carrier-Based Microwave Interferometry Sensing[J]. Sensors, 2020, 20(11)3079-3094.
[0012] [8]Wei T, Huang J, Lan XW, et al. Optical fiber sensor based on a radiofrequency Mach-Zehnder interferometer[J]. Optics Letter, 2012, 37(4): 647-649. Summary of the invention
[0013] The object of the present invention is to provide a multi-point maximum value demodulation algorithm based on microwave photon filter sensing, which includes a multi-point maximum value demodulation algorithm based on microwave photon filter sensing and a multi-point minimum value demodulation algorithm based on microwave photon filter sensing.
[0014] The frequency sweep curve containing each temperature and stress information to be measured is denoted as P(f(l))(1≤l≤L), where L is the total number of sampling points of each frequency sweep curve, and the lth frequency can be expressed as:
[0015] f(l)=f0+(l-1)Δf(1≤l≤L) (5)
[0016] The radio frequency power P(l) corresponding to the lth frequency. In order to accurately represent the collected radio frequency power, a row vector P = [P(1), ...P(l), ...P(L)] can be used.
[0017] A multi-point maximum demodulation algorithm based on microwave photon filter sensing is to calculate the sampling points where the first n RF amplitudes in the swept frequency data containing the temperature and stress information to be measured obtain the maximum value:
[0018] For each P, reorder the elements from large to small according to their size, denoted as P max =[P(f1), P(f2), ..., P(f l ), ..., P(f L )] and record the frequencies of the elements from large to small. These frequencies constitute the row vector F max =[f1, f2, ..., f l , ..., f L], and satisfy: P(f1)>P(f2)>…P(f l )>…P(f L ).
[0019] For each RF power vector P of the RF curve of the frequency sweep, the following multi-point maximum demodulation algorithm is used to extract the maximum RF resonance frequency estimate: as follows:
[0020] n is any integer and n<L. (6)
[0021] A multi-point minimum demodulation algorithm based on microwave photon filter sensing is to calculate the sampling points where the first n RF amplitudes in the swept frequency data containing the temperature and stress information to be measured obtain the minimum value:
[0022] For each P, reorder the elements from small to large according to their size, denoted as P min =[P(f1), P(f2), ..., P(f l ), ..., P(f L )] and record the frequencies of the elements from small to large. These frequencies constitute the row vector F min =[f1, f2, ..., f l , ..., f L ], and satisfy: {(f1)<P(f2)<...<P(f l )<...<P(f L ).
[0023] For each RF power vector P of the RF curve of the frequency sweep, the following multi-point minimum demodulation algorithm is used to extract the minimum RF resonant frequency estimate: as follows:
[0024] n is any integer and n<L. (7)
[0025] More specifically, the value of the number of sampling points n used for calculation of the multi-point maximum demodulation algorithm based on the microwave photon filter sensor system can be flexibly selected according to actual conditions.
[0026] It should be understood that the foregoing general description and the following detailed description are exemplary illustrations and explanations and should not be used as limitations on the content of the invention. BRIEF DESCRIPTION OF THE DRAWINGS
[0027] The advantages and features of the present invention will be clearly illustrated through the specific embodiments of the present invention in conjunction with the accompanying drawings.
[0028] Figure 1This is a flow chart of signal demodulation using a multi-point maximum demodulation algorithm in a microwave photon filter-based sensing system.
[0029] Figure 2 The structure of the microwave photon filter sensing system suitable for the multi-point maximum demodulation algorithm, (a) is the unidirectional fiber ring interferometer structure, and (b) is the Mach-Zehnder interferometer structure.
[0030] Figure 3 The figure is a schematic diagram of a microwave photon filter temperature sensing system device based on an optical fiber unidirectional ring structure.
[0031] Figure 4 This is a fitting result diagram obtained by extracting the maximum RF resonance frequency using two multi-point maximum demodulation algorithms, the three-point maximum method and the five-point maximum method, and the traditional maximum discrimination method.
[0032] Figure 5 This is a fitting result diagram obtained by extracting the minimum RF resonant frequency using two multi-point minimum demodulation algorithms, the three-point minimum method and the five-point minimum method, and the traditional minimum judgment method. DETAILED DESCRIPTION
[0033] Figure 1 A flowchart of using a multi-point maximum demodulation algorithm for signal demodulation in a microwave photon filter-based sensor system, including (a) a multi-point maximum demodulation algorithm based on microwave photon filter sensing and (b) a multi-point minimum demodulation algorithm based on microwave photon filter sensing.
[0034] A multi-point maximum demodulation algorithm based on microwave photon filter sensing:
[0035] a) The light source emits a light signal of a certain power;
[0036] b) modulating the optical signal emitted by the light source through the modulation part;
[0037] c) The modulated optical signal enters the interferometer part to form a microwave photon filter;
[0038] d) performing photoelectric conversion on the optical signal loaded with the microwave signal through an optical receiver;
[0039] e) The analyzer collects raw frequency sweep data including temperature and stress to be measured
[0040] f) using a multi-point maximum value method to process the sampling points corresponding to the first n maximum values of the RF amplitude for the frequency sweep data containing temperature and stress to be measured, and obtaining the maximum RF resonance frequency;
[0041] g) Extract the data of the maximum radio frequency resonance frequency changing with the temperature and stress to be measured, and form a database of temperature or stress sensing.
[0042] A multi-point minimum demodulation algorithm based on microwave photon filter sensing:
[0043] a) The light source emits a light signal of a certain power;
[0044] b) modulating the optical signal emitted by the light source through the modulation part;
[0045] c) The modulated optical signal enters the interferometer part to form a microwave photon filter;
[0046] d) performing photoelectric conversion on the optical signal loaded with the microwave signal through an optical receiver;
[0047] e) The analyzer collects raw frequency sweep data including temperature and stress to be measured
[0048] f) using a multi-point minimum value method to process the sampling points corresponding to the first n minimum values of the RF amplitude for the frequency sweep data containing temperature and stress to be measured, and obtaining the minimum RF resonance frequency;
[0049] g) Extract the data of the minimum RF resonance frequency changing with the temperature and stress to be measured, and form a database of temperature or stress sensing.
[0050] More specifically, for each temperature and stress, steps a) to f) are used to extract the information of the maximum frequency changing with the measured value, and finally form a sensor database for the measured value.
[0051] Figure 2 The structure of the microwave photon filter sensing system suitable for the multi-point maximum demodulation algorithm includes:
[0052] Light source (101): used to generate the light signal required by the light sensing system;
[0053] More specifically, in the microwave photon filter system, the light source (101) can be divided into two types: a coherent light source and an incoherent light source.
[0054] More specifically, if the light waves continuously emitted by a light source can be superimposed to produce interference, it is a coherent light source, otherwise it is an incoherent light source.
[0055] The modulation part (102) comprises an electro-optical intensity modulator, which is used to modulate the microwave signal within the frequency range [f0, f0+(L-1)Δf] onto the optical signal emitted by the light source (101) in a linear frequency sweeping manner;
[0056] More specifically, the sweep time of the entire RF frequency range is T s, the interval between the two sampling frequencies is T s / L;
[0057] Interferometer (103): used to form a microwave photon filter;
[0058] More specifically, two common interferometer structures include a fiber optic one-way ring interferometer (201) and a Mach-Zehnder interferometer (202);
[0059] More specifically, the fiber optic unidirectional ring interferometer (201) is composed of a 2x2 fiber optic coupler (301) as a beam splitter and a section of transmission fiber 1, wherein the fiber optic coupler (301) comprises two input ports a and b and two output ports c and d, the transmission fiber 1 comprises port a and port b, the input port a or input port b of the fiber optic coupler (301) is connected to the port a or port b of the transmission fiber optic 1, and the port b or port a of the transmission fiber optic 1 is connected to the output port c or output port d of the fiber optic coupler (301), forming a unidirectional fiber optic ring structure;
[0060] More specifically, the Mach-Zehnder interferometer (202) is composed of two 2x2 fiber couplers as beam splitters, a transmission fiber 2 and a reference fiber 1: a fiber coupler (302) comprising two input ports a1 and b1, and two output ports c1 and d1, wherein a1 and b1 serve as the a port and b port of the Mach-Zehnder interferometer (202) respectively, and c1 and d1 are connected to the transmission fiber 2 and the reference fiber 1 respectively, or c1 and d1 are connected to the reference fiber 1 and the transmission fiber 2 respectively; a fiber coupler (303) comprising two input optical ports a2 and b2, and two output optical ports c2 and d2, wherein c2 and d2 serve as the c port and d port of the Mach-Zehnder interferometer (202) respectively, and a2 and b2 are connected to the transmission fiber 2 and the reference fiber 1 respectively, or a2 and b2 are connected to the reference fiber 1 and the transmission fiber 2 respectively;
[0061] More specifically, the transmission optical fiber 2 has a length of L1 and the reference optical fiber 1 has a length of L2, and both are placed in a place where the temperature and stress are relatively stable to form microwave photon interference.
[0062] More specifically, the interferometer (103) includes not only components forming a microwave photon filter but also a sensing element.
[0063] More specifically, the sensing element should be placed inside the ring of a fiber optic one-way ring interferometer (201) or in one arm of a Mach-Zehnder interferometer (202).
[0064] More specifically, the change of the sensing information causes the change of the signal of the microwave photon interference, thereby changing the response curve of the microwave photon filter collected.
[0065] An optical receiver (104) is used to perform photoelectric conversion on an optical signal loaded with a microwave signal;
[0066] More specifically, the optical receiver (104) generally uses a photodetector to perform photoelectric conversion.
[0067] Analyzer (105): used to measure the radio frequency signal received by the optical receiver (104) and collect the radio frequency power at each frequency point.
[0068] More specifically, the analyzer (105) generally uses a spectrum analyzer and a vector network analyzer to measure and collect signals.
[0069] The multi-point maximum demodulation algorithm is adopted in the microwave photon filter sensing system according to the present invention to conduct verification and analysis through the following experiments:
[0070] Figure 3 The figure is a schematic diagram of a microwave photon filter temperature sensing system device based on a fiber ring structure. The spontaneously radiated broadband optical signal is emitted by an erbium-doped fiber amplifier (EDFA). The electro-optic modulator (EOM) modulates the sine wave signal generated by the RF signal source on the broadband optical signal. The modulated optical signal is transmitted to the fiber ring, which is a unidirectional fiber ring composed of an optical coupler (OC) and a sensing fiber. In the unidirectional fiber ring interferometer, the signal enters the optical coupler from the input port IN1 and is divided into two signals. One signal is directly emitted to the photodetector (PD) through the output port OUT1, and the other signal is transmitted through the sensing fiber through the output port OUT2, reaches the input port IN2, and enters the OC and is divided into two signals. One signal enters the unidirectional fiber ring again, and the other signal has a time delay difference with the previous signal in the OC, forming a microwave photon filter (MPF). The signal circulates repeatedly in the fiber ring to infinity. After sampling the fiber ring infinitely, the signal is converted into electrical and optical signals at the PD and collected by the electrical spectrum analysis function of an electrical spectrum analyzer (ESA, model Tektronix MDO3052).
[0071] The output optical power of EDFA was set to 12.0 dBm, the frequency range of the modulated RF signal was 137.2 MHz-138.5 MHz, the amplitude was 4 Vpp, the bias voltage of EOM was 1 V, the scanning time was 180 s, the scanning bandwidth of ESA was set to 2 kHz, the experimental temperature range was selected to be 25 C-65 C, and 5 sets of raw frequency sweep data were collected using a spectrum analyzer at each temperature.
[0072] The three-point maximum method and five-point maximum method in the multi-point maximum demodulation algorithm and the traditional maximum discrimination method are used to extract the maximum RF resonance frequency of N groups of experimental data and perform polynomial first-order fitting to obtain Figure 4Fitting results, where (a) is the first-order fitting degree and (b) is the first-order fitting coefficient.
[0073] The three-point minimum method and five-point minimum method in the multi-point minimum method and the traditional minimum discrimination method are used to extract the minimum RF resonance frequency of the M group of experimental data and perform polynomial fitting to obtain Figure 5 Fitting results, where (a) is the first-order fitting degree and (b) is the first-order fitting coefficient.
[0074] Depend on Figure 4 It can be seen that the system fit and sensitivity extracted by the maximum discrimination method have large fluctuations. This is mainly due to the noise in the sensor system, which makes it impossible to directly extract temperature and stress sensing information through the drift of the maximum RF resonance frequency. The three-point maximum method has a significant improvement in the fit. For data points that are not significantly improved by the three-point maximum method, the five-point maximum method can further improve them. According to calculations, the five-point maximum method can increase the system fit corresponding to the extracted maximum RF resonance frequency by 1.78 times and the sensitivity by 1.10 times.
[0075] Similarly, due to the presence of noise in the sensing system, it is impossible to use the minimum value discrimination method to directly extract the temperature and stress sensing information through the drift of the minimum RF resonant frequency. Figure 5 It can be seen that the system fit and sensitivity extracted by the minimum value discrimination method also have large fluctuations, and the five-point minimum value method has a significant improvement in both system fit and sensitivity. According to calculations, the five-point minimum value method can increase the system fit corresponding to the extracted minimum RF resonance frequency to 1.07 times and the sensitivity to 1.18 times.
[0076] Other embodiments of the present invention are readily apparent to those skilled in the art. The description and embodiments are to be considered as exemplary only, with the true scope and spirit of the present invention being defined by the claims.
Claims
1. A multi-point maximum demodulation algorithm based on microwave photon filter sensing, characterized by: This method calculates the first n sampling points where the RF amplitude obtains the maximum value in the swept frequency data containing some information to be measured, such as temperature and stress. The swept frequency curve containing each temperature and stress information to be measured is recorded as P(f(l))(1≤l≤L), L is the total number of sampling points of each swept frequency curve, and the lth frequency is expressed as: f(l)=f0+(l-1)Δf(1≤l≤L), (1) The RF power P(l) corresponding to the lth frequency, f0 is the first frequency of the sweep, f(L) is the Lth frequency of the sweep, Δf = (f(L)-f0) / (L-1), in order to accurately represent the collected RF power, a row vector P = [P(1), ...P(l), ...P(L)] is used to represent it; for each row vector P, the elements are reordered from large to small according to their size, and are recorded as P max =[P(f1), P(f2), ..., P(f l ), ..., P(f L )] and record the frequencies of the elements from large to small. These frequencies constitute the row vector F max =[f1, f2, ..., f l , ..., f L ], and satisfy: P(f1)>P(f2)>…P(f l )>…P(f L ); For each RF power vector P of the RF curve of the frequency sweep, the following multi-point maximum demodulation algorithm is used to extract the maximum RF resonance frequency estimate: as follows: n is any integer greater than 2 and n <L。 (2) 2. A multi-point minimum demodulation algorithm based on microwave photon filter sensing, characterized by: This method calculates the first n sampling points where the RF amplitude obtains the maximum value in the swept frequency data containing some information to be measured, such as temperature and stress. The swept frequency curve containing each temperature and stress information to be measured is recorded as P(f(l))(1≤l≤L), L is the total number of sampling points of each swept frequency curve, and the lth frequency is expressed as: f(l)=f0+(l-1)Δf(1≤l≤L), (3) The RF power P(l) corresponding to the lth frequency; f0 is the first frequency of the sweep, f(L) is the Lth frequency of the sweep, Δf = (f(L)-f0) / (L-1), in order to accurately represent the collected RF power, a row vector P = [P(1), ...P(l), ...P(L)] is used to represent it; for each P, the elements are reordered from small to large according to their size, and recorded as P min =[P(f1), P(f2), ..., P(f l ), ..., P(f L )] and record the frequencies of the elements from small to large. These frequencies constitute the row vector F min =[f1, f2, ..., f l , ..., f L ], and satisfy: P(f1) <P(f2)<…<P(f l )<… <P(f L ); For each frequency sweep of the RF power vector P of the RF curve, the following multi-point maximum demodulation algorithm is used to extract the minimum RF resonance frequency estimate: as follows: n is any integer greater than 2 and n <L。 (4) 3. The multi-point maximum demodulation algorithm or the multi-point minimum demodulation algorithm as claimed in claim 1 or claim 2, the method is based on the following microwave photon filter sensor system, the structure of which includes: Light source (101): used to generate the light signal required by the light sensing system; The modulation part (102) includes an electro-optical intensity modulator, which is used to modulate the microwave signal in the frequency range [f0, f0+(L-1)Δf] onto the optical signal emitted by the light source (101) in a linear frequency sweeping manner, wherein the frequency sweeping time of the entire RF frequency range is T s , the interval between the two sampling frequencies is T s / L; Interferometer (103): used to form a microwave photon filter; An optical receiver (104) is used to perform photoelectric conversion on an optical signal loaded with a microwave signal; Analyzer (105): used to measure the radio frequency signal received by the optical receiver (104) and collect the radio frequency power at each frequency point.
4. The multi-point maximum demodulation algorithm based on microwave photon filter sensing as claimed in claim 1 can determine the maximum radio frequency resonance frequency of the microwave photon filter sensing system under each condition to be measured according to the following method: a) The light source emits a light signal of a certain power; b) modulating the optical signal emitted by the light source through the modulation part; c) The modulated optical signal enters the interferometer part to form a microwave photon filter; d) performing photoelectric conversion on the optical signal loaded with the microwave signal through an optical receiver; e) The analyzer collects raw frequency sweep data including temperature and stress to be measured f) using a multi-point maximum value method to process the sampling points corresponding to the first n maximum values of the RF amplitude for the frequency sweep data containing temperature and stress to be measured, and obtaining the maximum RF resonance frequency; g) Extract the data of the maximum radio frequency resonance frequency changing with the temperature and stress to be measured, and form a database of temperature or stress sensing.
5. The multi-point minimum demodulation algorithm based on microwave photon filter sensing as claimed in claim 2 can determine the minimum radio frequency resonance frequency of the microwave photon filter sensing system under each condition to be measured according to the following method: a) The light source emits a light signal of a certain power; b) modulating the optical signal emitted by the light source through the modulation part; c) The modulated optical signal enters the interferometer part to form a microwave photon filter; d) performing photoelectric conversion on the optical signal loaded with the microwave signal through an optical receiver; e) The analyzer collects raw frequency sweep data including temperature and stress to be measured f) using a multi-point minimum value method to process the sampling points corresponding to the first n minimum values of the RF amplitude for the frequency sweep data containing temperature and stress to be measured, and obtaining the minimum RF resonance frequency; g) Extract the data of the minimum RF resonance frequency changing with the temperature and stress to be measured, and form a database of temperature or stress sensing.
6. The multi-point maximum demodulation algorithm or the multi-point minimum demodulation algorithm as claimed in claim 3, characterized in that: The interferometer (103) is a fiber optic one-way ring interferometer (201) formed by connecting an output port and an input port of a 2×2 fiber optic coupler (301) serving as a beam splitter.
7. The multi-point maximum demodulation algorithm or the multi-point minimum demodulation algorithm as claimed in claim 3, characterized in that: The interferometer (103) is a Mach-Zehnder interferometer (202) composed of two 2×2 fiber couplers (301) as beam splitters, a transmission fiber and a reference fiber.
