Design of Fiber Bragg Grating Demodulator Based on Single Chip Microcomputer

The application of fiber Bragg grating sensors is a burgeoning field with a very broad development prospect. At present, the main obstacle that limits the large-scale practical application of fiber Bragg grating sensors is the demodulation of the sensing signal. There are many fiber Bragg grating sensor demodulation methods, but there are not many demodulation products that can be used in practice, and they are expensive. Therefore, researching and developing demodulation systems suitable for practical engineering applications and reducing the cost of demodulation systems are key issues to promote fiber Bragg grating sensors in practical engineering applications.

In view of this, in order to meet the needs of engineering applications, this paper proposes a fiber Bragg grating demodulation technology based on a single-chip microcomputer, that is, using the currently widely used and relatively cheap single-chip microcomputer as the MCU for signal acquisition and processing, to develop a demodulator that is high-precision, cheap, portable, can perform fast measurements, and can easily obtain the size of the measured parameter. In order to solve the problem of the slow speed of a single single-chip microcomputer, a dual CPU is used in the system, in which one single-chip microcomputer completes the signal demodulation algorithm, while the other single-chip microcomputer completes the logic control, human-machine interface and communication with the host computer, and realizes dual-machine data sharing through dual-port RAM.

1. Demodulation system structure and principle

The overall structure of the demodulation system is shown in Figure 1. It mainly consists of three parts: Bragg grating (measurement grating), fiber Bragg demodulator, and computer. The fiber Bragg demodulator can be divided into two parts, the analog circuit part and the digital circuit part. The function of the analog circuit part is to convert the strain or temperature change of the Bragg grating (measurement grating) into a corresponding electrical signal, and the digital part converts the electrical signal into a digital signal that can be directly used by the host computer, which can be a wavelength value or a temperature or strain value. The MCU used to realize this function is a single-chip microcomputer.

Figure 1 Overall structure of demodulation system

The demodulation principle of the demodulation system is based on the working principle of the tunable Fabry-Perot cavity (FP demodulation). The fiber FP cavity filter used for Bragg grating sensor signal demodulation is actually a voltage-controlled optical bandpass filter, and piezoelectric ceramics are usually used as the driving element for the change of the FP cavity length. When a scanning voltage is applied to the piezoelectric ceramic, the piezoelectric ceramic will expand and contract, thereby changing the cavity length of the FP cavity and changing the wavelength of the light passing through the FP cavity. The intensity of the transmitted light is detected by the detector. When the detector detects the maximum light intensity, the voltage applied to the piezoelectric ceramic corresponds to the reflection wavelength of the FBG. In this way, an optical signal is injected into the Bragg fiber grating sensor, and the light reflected from the FBG sensor is added to the input end of the fiber FP cavity filter. By adding a triangular scanning voltage to the voltage-controlled end of the fiber FP cavity filter, a time domain electrical signal corresponding to the input light spectrum can be obtained at the output end of the fiber FP cavity filter. These time domain signals are shaped by the amplification circuit and the comparison circuit to obtain a series of pulse signals. We add some standard pulse signals with fixed wavelengths and positions to these pulse signals, and the relative position of each pulse in these pulse signals to the standard pulse contains the spectrum information of the reflected light of the FBG sensor. Figure 2 shows this demodulation process. Finally, the obtained pulses are converted into wavelength values ​​through the circuit composed of a single chip.

2. The composition and working method of the single-chip microcomputer demodulation system

The primary purpose of the microcontroller demodulation system is to process these pulse signals into corresponding wavelength values. Through the demodulation of the analog part, we get a pulse signal containing the relative position of the measurement grating and the standard grating in the scanning cycle. The standard grating corresponds to a fixed wavelength, and the position of its corresponding pulse signal in each scanning cycle is fixed (the standard grating uses a constant temperature circuit to keep the wavelength constant). If the relative position value of each pulse signal can be obtained, the wavelength value of the measurement grating can be obtained through the interpolation algorithm.

Figure 2 Fiber Bragg grating sensor signal demodulation process

In this demodulation system, the fiber Bragg grating produced by Wuhan University of Technology Optical Technology Co., Ltd. is used as the measuring grating, and the wavelength selector based on the FP cavity principle is used as the demodulation cavity. The measurement range can reach 30nm, the period of the triangular wave scanning signal is 1s, and the measurement frequency is 1Hz. The rising edge of the triangular wave scanning signal is divided into a finite number of counting points that can achieve the design accuracy, so that the position value of the FBG1, FBG2, .....FBGn grating array and the standard grating pulse signal in the rising edge of the triangular wave can be read out by a single-chip microcomputer. The function of another single-chip microcomputer is to use these values ​​to calculate the wavelength and communicate data with the computer. The circuit diagram is shown in the figure. The single-chip microcomputer here is 89C52, and 4060 is used to generate a stable counting pulse. When the triangular wave starts, the No. 1 single-chip microcomputer counts. When a pulse arrives, the counter value is recorded and stored in the on-chip RAM; when the triangular wave reaches the highest point, the counter is cleared, the position value is sent to the dual-port RAM, and then waits for the next count. When CPU1 starts counting, CPU2 takes data out of the dual-port RAM, calculates the wavelength value or temperature value corresponding to the pulse through interpolation or other algorithms, and communicates with the computer.

Circuit diagram

We can input more measurement pulses simultaneously through other I/O ports of the microcontroller. By improving the optical path and analog circuit, we can make 2-channel and 4-channel fiber Bragg grating demodulators to provide more measurement points, and the digital circuit does not need to be changed at all, only the software part needs to be adjusted.

3. System analysis and data processing

If the microcontroller needs to record each pulse completely and correctly, then its counting and transmission instructions must be completed within the pulse width of each pulse. If the pulse width is only 1 counting unit, that is, the counting and transmission instructions need to be completed within about 10 microseconds. The maximum operating frequency of AMTEL's 89C52 can reach 24MHz, and its clock cycle is 0.5 microseconds. As long as the counting and transmission instruction cycle does not exceed 20 clock cycles, the requirement can be met. A reasonable read and write program can obviously meet this requirement. Usually, the width of the pulse is generally much larger than 1 counting unit, so the change of the pulse can be recorded in real time. At the same time, the No. 2 microcontroller has 1s to take the data out of the RAM, calculate the median of the pulse, and then perform interpolation calculations. The time is also sufficient. If the algorithm is too complicated, such as using the Lagrange algorithm, etc., the position value can also be transmitted to the computer for data processing.

In the process of transmitting data from the microcontroller to the computer, errors may occur in the data. Error correction processing must be added to the communication program. Parity check methods can be used, such as single-byte check or multiple-byte check, etc. At the same time, in order to prevent occasional mutations in the grating position value, it is necessary to smooth the position value. Through the above processing methods, the computer can obtain a set of correct and stable data. In order to reduce the drift of the FP cavity and the influence of the system nonlinearity on the position value, we use a standard grating to compare and calculate with the measured grating, and a linear algorithm can be used for calculation. However, in actual application, it is found that when the grating to be measured is closer to the standard grating, the measurement value is more accurate; the farther away, the error is relatively large. In order to further improve the accuracy, 2 standards, 5 standards or comb filters can be used for piecewise linear interpolation calculations, which can greatly improve the measurement accuracy. Of course, more complex methods such as Lagrangian algorithm or multi-term formula can also be used to calculate the wavelength. In our instrument, the Lagrangian algorithm of 5 standard gratings is used to calculate the wavelength, and the temperature measurement accuracy can reach ±1℃.

4. Conclusion

The relationship between the relative position value of the pulse and the wavelength cannot be derived from theoretical knowledge at present, but it can be found through experiments and mathematical statistics to find out the change law and the corresponding relationship between them. Using this corresponding relationship, the relevant data processing is carried out in the single-chip microcomputer to obtain the magnitude of the measured temperature or stress. At present, we use the Lagrange algorithm and some appropriate data processing and calibration methods. Judging from the current working conditions of the demodulator, the effect is still good, the measurement accuracy can reach ±5pm, and the maximum repeatability error is 8pm. In order to increase the working frequency of the demodulator and improve the applicability of the demodulator, a demodulation circuit based on DSP or DSP+ARM can also be used, but the cost is relatively high.

FBG gratings have broad application prospects and can play an important role in communications, construction, machinery, medical treatment, aerospace, navigation, and mining. Theoretical research on FBG gratings has made great achievements so far. By adopting appropriate demodulation technology and reducing the cost of fiber Bragg grating, fiber Bragg grating sensors can be widely used in practical engineering.

The author's innovation: The fiber Bragg grating demodulator based on single-chip microcomputer is a demodulation system suitable for practical engineering applications, which greatly reduces the cost of the fiber Bragg grating demodulation system, facilitates its use in industrial sites, and enables the fiber Bragg grating sensor to be rapidly promoted in practical engineering applications.