Signal-to-noise ratio enhancement method, balance detection method and circuit

Abstract

The invention discloses a signal-to-noise ratio enhancement method, a balance detection method and a circuit, and the method comprises the steps: adjusting the gain and reference threshold voltage according to an output signal of an analog-to-digital converter, enabling the mean value of the threshold voltage to be closest to the mean value of a signal after gain processing, enabling the peak-to-peak value of the signal after gain processing to be larger than a threshold voltage difference value, and achieving the balance detection of the signal-to-noise ratio. Under the condition, noise carried in the signal after gain processing is used for driving comparison output to jump between a high level and a low level, a stochastic resonance effect is formed, and therefore the signal-to-noise ratio of the stochastic resonance signal is greatly increased.

Description

Technical Field

[0001] This invention relates to a signal-to-noise ratio enhancement method, a balanced detection method, and a circuit, belonging to the field of coherent wind lidar. Background Technology

[0002] In the field of coherent wind lidar, a balanced detector is usually used to convert the optical signal into an electrical signal. Specifically, the wind laser signal is converted into a wind electrical signal by a photodetector with a cathode and an anode connected together. The electrical signal is then amplified and filtered by a module (i.e., amplified and filtered in sequence) to obtain a wind laser radio frequency signal with a frequency in the MHz range. The radio frequency signal is then sampled by an analog-to-digital converter to obtain sampled data, and then a processor is used to analyze and process the sampled data.

[0003] Because wind measurement signals are extremely weak, the radio frequency (RF) signals obtained by balanced detection techniques are often submerged in noise, resulting in a low signal-to-noise ratio (SNR). To improve the detection range of coherent wind lidar, enhancing the SNR before sampling the RF signal is a better approach. Traditional enhancement methods involve filtering the RF signal to remove noise outside the frequency range of interest; however, this method has limited effectiveness in improving the SNR. Summary of the Invention

[0004] This invention provides a signal-to-noise ratio enhancement method, a balanced detection method, and a circuit, which solve the problems disclosed in the background art.

[0005] According to one aspect of this application, a signal-to-noise ratio enhancement method is provided, comprising: Gain processing is performed on the wind measurement laser radio frequency signal at the current moment; A new threshold voltage is generated based on the reference threshold voltage, the signal-to-noise ratio enhanced random resonance signal obtained at the previous moment, and a preset fixed relationship. The signal after gain processing at the current moment is compared with the new threshold voltage to obtain the random resonance signal with enhanced signal-to-noise ratio at the current moment. Among them, the random resonance signal is input to the analog-to-digital converter of the balanced detector, the wind measurement laser radio frequency signal is the output signal of the amplification and filtering module of the balanced detector; the new threshold voltage is either a high threshold voltage or a low threshold voltage. The gain and reference threshold voltage are adjusted based on the analog-to-digital converter output signal. The goal of adjusting the reference threshold voltage is to make the average threshold voltage as close as possible to the average signal after gain processing. The goal of adjusting the gain is to make the peak-to-peak value of the signal after gain processing greater than the threshold voltage difference. The average threshold voltage is the average of the high threshold voltage and the low threshold voltage, and the threshold voltage difference is the difference between the high threshold voltage and the low threshold voltage.

[0006] Furthermore, the goal of adjusting the gain also includes ensuring that the ratio between the root mean square of the noise in the signal after gain processing and the difference in the threshold voltage is within the range of a pre-calibrated optimal ratio.

[0007] Furthermore, adjusting the gain based on the analog-to-digital converter output signal includes: The gain is increased if the switching frequency of the analog-to-digital converter output signal is lower than the pre-calibrated optimal frequency range, and decreased if the switching frequency is higher than the pre-calibrated optimal frequency range. If the switching frequency is within the pre-calibrated optimal frequency range, the gain remains unchanged.

[0008] Further, adjusting the reference threshold voltage based on the analog-to-digital converter output signal includes: Calculate the mean value of the analog-to-digital converter output signal. If the mean value is greater than 0.5, decrease the reference threshold voltage; if the mean value is less than 0.5, increase the reference threshold voltage.

[0009] According to another aspect of this application, a signal-to-noise ratio enhancement circuit is provided, including a gain module and a Schmitt trigger; The input terminal of the gain module is connected to the amplification and filtering module of the balanced detector. The first and second input terminals of the Schmitt trigger are connected to the output terminal of the gain module and the processor of the balanced detector, respectively. The output terminal of the Schmitt trigger is connected to the analog-to-digital converter of the balanced detector. The gain module is used to perform gain processing on the wind measurement laser radio frequency signal at the current moment; A Schmitt trigger is used to generate a new threshold voltage based on a reference threshold voltage, the signal-to-noise ratio enhanced random resonance signal obtained at the previous time step, and a preset fixed relationship; the signal after gain processing at the current time step is compared with the new threshold voltage to obtain the signal-to-noise ratio enhanced random resonance signal at the current time step; wherein, the new threshold voltage is a high threshold voltage or a low threshold voltage. The processor generates a gain and a reference threshold voltage based on the output signal of the analog-to-digital converter, and sends the generated gain and reference threshold voltage to the gain module and the Schmitt trigger, respectively. The processor generates a reference threshold voltage with the objective of making the mean threshold voltage as close as possible to the mean signal after gain processing. The processor generates a gain with the objective of making the peak-to-peak value of the signal after gain processing greater than the threshold voltage difference. The mean threshold voltage is the average of the high threshold voltage and the low threshold voltage, and the threshold voltage difference is the difference between the high threshold voltage and the low threshold voltage.

[0010] Furthermore, the processor's goal in generating gain also includes ensuring that the ratio between the root mean square of the noise in the signal after gain processing and the threshold voltage difference is within a pre-calibrated optimal ratio range.

[0011] Furthermore, in the processor, a gain is generated based on the analog-to-digital converter output signal, including: The gain is increased if the switching frequency of the analog-to-digital converter output signal is lower than the pre-calibrated optimal frequency range, and decreased if the switching frequency is higher than the pre-calibrated optimal frequency range. If the switching frequency is within the pre-calibrated optimal frequency range, the gain remains unchanged.

[0012] Furthermore, in the processor, a reference threshold voltage is generated based on the analog-to-digital converter output signal, including: Calculate the mean value of the analog-to-digital converter output signal. If the mean value is greater than 0.5, decrease the reference threshold voltage; if the mean value is less than 0.5, increase the reference threshold voltage.

[0013] According to another aspect of this application, a balanced detection method is provided, including the above-described signal-to-noise ratio enhancement method.

[0014] According to another aspect of this application, a balanced detection circuit is provided, including the above-described signal-to-noise ratio enhancement circuit.

[0015] The beneficial effects achieved by this invention are as follows: This invention adjusts the gain and reference threshold voltage according to the output signal of the analog-to-digital converter, so that the average value of the threshold voltage is closest to the average value of the signal after gain processing, and the peak-to-peak value of the signal after gain processing is greater than the difference in threshold voltage. In this case, the noise carried in the signal after gain processing is used to drive the output of the comparison to jump between high and low levels, forming a random resonance effect, thereby greatly enhancing the signal-to-noise ratio of the random resonance signal. Attached Figure Description

[0016] Figure 1 The flowchart shows the signal-to-noise ratio enhancement method. Figure 2 This is a schematic diagram of the first type of signal-to-noise ratio enhancement circuit. Figure 3 The second schematic diagram of the signal-to-noise ratio enhancement circuit. Figure 4 This is a schematic diagram of a balanced detection circuit. Detailed Implementation

[0017] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit this application or its application or use. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0018] Unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps described in these embodiments do not limit the scope of this application.

[0019] At the same time, it should be understood that, for ease of description, the dimensions of the various parts shown in the accompanying drawings are not drawn according to actual scale.

[0020] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and equipment should be considered part of the specification.

[0021] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0022] It should be noted that similar symbols and letters in the following figures represent similar items; therefore, once an item is defined in one figure, it does not need to be discussed further in subsequent figures.

[0023] Furthermore, in the description of the embodiments of this application, the terms "first," "second," etc., are used only for distinguishing descriptions and should not be construed as indicating or implying relative importance. Therefore, features defined with "first" or "second" may explicitly or implicitly include one or more features.

[0024] See Figure 1 , Figure 1 This is a flowchart of a signal-to-noise ratio enhancement method provided in an embodiment of this application. The method may include at least the following steps: Step 1: Gain processing is performed on the wind measurement laser radio frequency signal; wherein, the wind measurement laser radio frequency signal is the output signal of the amplification and filtering module of the balanced detector.

[0025] It should be noted that gain processing of the wind measurement laser RF signal mainly involves amplifying or attenuating the signal based on gain. The signal after gain processing can be denoted as V. VGA The purpose of amplifying the signal here is to create a stochastic resonance effect, which in turn enhances the signal-to-noise ratio.

[0026] Step 2: Based on the reference threshold voltage, the signal-to-noise ratio enhanced random resonance signal obtained at the previous moment, and a preset fixed relationship, a new threshold voltage is generated. The signal after gain processing at the current moment is compared with the new threshold voltage to obtain the signal-to-noise ratio enhanced random resonance signal at the current moment. The random resonance signal is input to the analog-to-digital converter of the balanced detector. The new threshold voltage is either a high threshold voltage or a low threshold voltage. The gain and reference threshold voltage are adjusted according to the output signal of the analog-to-digital converter. The goal of adjusting the reference threshold voltage is to make the average threshold voltage as close as possible to the average value of the signal after gain processing. The goal of adjusting the gain is to make the peak-to-peak value of the signal after gain processing greater than the threshold voltage difference. The average threshold voltage is the average of the high threshold voltage and the low threshold voltage, and the threshold voltage difference is the difference between the high threshold voltage and the low threshold voltage.

[0027] It should be noted that, for ease of description, the reference threshold voltage is denoted as V. ref Using the signal-to-noise ratio enhanced random resonance signal obtained in the previous moment and a preset fixed relationship, V at different moments ref Split into high threshold voltage V H and low threshold voltage V L By adjusting V ref It can change V H and V L The mean, i.e. (V H +V L ) / 2, but the difference between the two (V) H -V L ) remains unchanged. When V VGA The voltage is higher than V H When the two are compared, the output is low; when V VGA The voltage is lower than V L When the two are compared, a high level is output. This high and low level constitutes a random resonant signal, which is used for sampling by the subsequent analog-to-digital converter.

[0028] To achieve the stochastic resonance effect, V needs to be adjusted. ref and V VGA Make appropriate adjustments, more specifically, based on the analog-to-digital converter output signal, to adjust the gain and reference threshold voltage appropriately, when adjusted to (V H +V L ) / 2 and V VGA The mean is closest to V VGA The peak-to-peak value is greater than (V H -V L When V VGA The noise carried in the signal is used to drive the output of the comparison to jump between high and low levels, forming a random resonance effect, thereby improving the signal-to-noise ratio of the signal fed into the analog-to-digital converter.

[0029] It should be noted that if V VGA If the root mean square of the noise is too small, then V VGA It is difficult to make the compared output change frequently, V VGA The signal carried in V is difficult to transmit into the random resonance signal. VGA If the root mean square of the noise is too large, the output jump after comparison will tend to be completely random, V VGA The signal carried in the signal is also difficult to transmit to the random resonance signal. Therefore, in some embodiments, in addition to considering the two adjustment objectives mentioned above, it is also necessary to further consider V. VGA The root mean square of noise and (V) H -V L Specifically, this involves adjusting the gain to make the processed signal V... VGA The root mean square difference between the noise level and the threshold voltage (V) H -V L Within the pre-defined optimal proportional relationship range.

[0030] Only when V VGA The root mean square of noise and (V) H -V L A random resonance effect can only be formed within a certain proportional range. At this point, the energy of the noise is used to help the signal cross the threshold, and the transition probability of the random resonance signal is affected by V. VGA To mitigate the impact of the signals carried in the medium and to maximize the signal-to-noise ratio.

[0031] It should be noted that the calibration of the optimal proportional relationship range can be as follows: A signal with a signal-to-noise ratio below 0 dB is generated using a noise source and a sinusoidal signal source as the wind measurement laser radio frequency signal. This signal is then used to calibrate the V... VGA With (V) H -V L When the ratio is within the optimal range, the random resonance signal is converted into a spectral signal through Fourier transform. The signal-to-noise ratio (SNR) at the corresponding frequency points can be observed; this is the SNR of the random resonance signal. Adjust the gain and plot "V". VGA With (V) H -V L The graph showing the relationship between the "ratio" and the signal-to-noise ratio of the random resonance signal is used to determine the optimal range of the ratio. Simultaneously, "V..." VGA With (V) H -V L A graph can also be plotted between the ratio and the jumping frequency of the random resonance signal, from which V can be obtained. VGA With (V) H -V L Within the optimal ratio range, the frequency range of the random resonance signal.

[0032] It should be noted that adjusting the gain means increasing or decreasing the value. Since the data output by the analog-to-digital converter is high or low level, when the random resonance effect reaches the strongest signal-to-noise ratio enhancement effect, the switching frequency of the analog-to-digital converter output signal will be within the optimal frequency range. This range can be determined by the calibration mentioned above.

[0033] Therefore, the specific process of gain adjustment can be as follows: Statistically analyze the switching frequency of the analog-to-digital converter output signal; if the switching frequency is lower than the pre-calibrated optimal frequency range, increase the gain, thereby increasing V. VGA If the voltage switching frequency is higher than the pre-calibrated optimal frequency range, reduce the gain, thereby reducing V. VGA If the voltage switching frequency is within the pre-calibrated optimal frequency range, the gain remains constant. Adjust V by adjusting the gain. VGA This allows the switching frequency to be controlled within the range where the random resonance effect is strongest.

[0034] It should be noted that the step size for increasing or decreasing can be selected manually. Generally, the smaller the step size, the slower the gain convergence speed. However, during the circuit operation, once a random resonance effect is formed, the adjustment required is not very large, so this step size is often selected to be relatively small.

[0035] It should be noted that V is adjusted in real time. VGA To achieve the strongest stochastic resonance effect, in order to avoid over- or under-controlling the gain by the balanced detector processor, PID (proportional-integral-derivative) control technology can be used to adjust V. VGA This allows for real-time tracking of the root mean square change of noise and rapid convergence to the optimal value.

[0036] It should be noted that, similarly, for V ref Adjustment involves increasing or decreasing the value; when (V) H +V L ) / 2 equals V VGA When comparing the mean of V, the quantization is... VGA The probability of outputting a high level and a low level is equal, therefore the probability of obtaining data 0 and data 1 after sampling is also equal. Therefore, in order to make V... VGA The mean should be as close as possible to (V) H +V L To achieve the strongest random resonance effect, the mean value of the analog-to-digital converter output signal can be calculated. If the mean value is greater than 0.5, the reference threshold voltage is decreased; if the mean value is less than 0.5, the reference threshold voltage is increased. The adjustment step of the reference threshold voltage can be determined based on the measured mean adjustment speed.

[0037] It should be noted that the gain and reference threshold voltage are adjusted simultaneously, which speeds up the convergence to the state of strongest stochastic resonance, making it suitable for implementation in digital chips such as FPGAs. Alternatively, the gain and reference threshold voltage can be adjusted alternately, which will result in a slower convergence speed, but is simpler to implement and suitable for chips such as microcontrollers.

[0038] The above method enhances the signal-to-noise ratio between the amplification and filtering module and the analog-to-digital converter in the existing balanced detection circuit. This method creates a random resonance effect, thereby greatly enhancing the signal-to-noise ratio of the random resonance signal.

[0039] The above method can be implemented programmatically, which can be integrated into the processor of the balancing detector. However, since the processor also needs to perform balancing detection analysis on the sampled data, and the processor's resources are limited, the above method is generally implemented using a circuit. For specific circuit details, please refer to [link to circuit implementation details]. Figure 2 The signal-to-noise ratio enhancement circuit may include at least a gain module and a Schmitt trigger; the input of the gain module is connected to the amplification and filtering module of the balanced detector, the first and second inputs of the Schmitt trigger are respectively connected to the output of the gain module and the processor of the balanced detector, and the output of the Schmitt trigger is connected to the analog-to-digital converter of the balanced detector.

[0040] The gain module is used to perform gain processing on the wind measurement laser RF signal at the current moment. Specifically, it amplifies or attenuates the wind measurement laser RF signal according to the gain. This module can use the programmable variable gain amplifier AD8370AREZ, whose gain adjustment can achieve a high dynamic range of 45 dB.

[0041] The Schmitt trigger is used to generate a new threshold voltage based on a reference threshold voltage, the signal-to-noise ratio enhanced random resonance signal obtained at the previous time step, and a preset fixed relationship; the signal after gain processing at the current time step is compared with the new threshold voltage to obtain the signal-to-noise ratio enhanced random resonance signal at the current time step; wherein, the new threshold voltage is either a high threshold voltage or a low threshold voltage.

[0042] A Schmitt trigger mainly includes a comparator, resistors R3 and R4. The comparator can be a TLV3603DCKR, with a switching frequency of up to 325 MHz. The inverting input of the comparator is connected to the output of the gain module, and the non-inverting input of the comparator is connected to the processor through resistor R3. The output of the comparator serves as the output of the Schmitt trigger, and this output is connected to the non-inverting input of the comparator through resistor R4. Resistor R3 can be 10 ohms, and resistor R4 can be 3.3K ohms.

[0043] It should be noted that the reference threshold voltage generated by the processor is generally a digital signal. If no digital-to-analog conversion is performed internally by the processor, see [link to relevant documentation]. Figure 3 A digital-to-analog converter (DAC) needs to be added to the circuit described above. Resistor R3 is connected to the processor through the DAC. The DAC mainly converts the reference threshold voltage into an analog signal. An AD9744ARUZRL7 with a working clock frequency of 100 MHz can be used.

[0044] Furthermore, in order to improve the driving capability of the digital-to-analog converter output signal and avoid the output voltage of the digital-to-analog converter from deviating significantly from the expected value due to insufficient driving capability, an amplifier can be added to the output terminal of the digital-to-analog converter. That is, resistor R3 is connected to the processor in sequence through the amplifier and the digital-to-analog converter. The amplifier can be an OPA820IDR, and the amplification gain can be set to 1.

[0045] Assume the amplified reference threshold voltage is V a If the comparator output voltage is Vs, then the voltage at the non-inverting input of the comparator is (Vs). a R4 + V s R3) / (R3+R4). Through resistor voltage division and output feedback, when the comparator outputs a high level, the voltage at the non-inverting input of the comparator is determined by V. a Adjusted to V H When the comparator outputs a low level, the voltage at the non-inverting input of the comparator is determined by V. a Adjusted to V L .

[0046] To enhance the signal-to-noise ratio, the processor of the balanced detector needs to be equipped with an adaptation function, namely, generating a gain and a reference threshold voltage based on the analog-to-digital converter output signal, and sending the generated gain and reference threshold voltage to the gain module and Schmitt trigger respectively (if using...). Figure 3 The structure of the reference threshold voltage is then sent to the digital-to-analog converter. The processor generates the reference threshold voltage with the goal of making the mean threshold voltage as close as possible to the mean signal after gain processing. The processor generates the gain with the goal of making the peak-to-peak value of the signal after gain processing greater than the threshold voltage difference. The mean threshold voltage is the average of the high threshold voltage and the low threshold voltage, and the threshold voltage difference is the difference between the high threshold voltage and the low threshold voltage.

[0047] The circuit adjusts the gain and reference threshold voltage based on the output signal of the analog-to-digital converter, so that the average threshold voltage in the Schmitt trigger is as close as possible to the average signal after gain processing, and the peak-to-peak value of the signal after gain processing is greater than the difference in threshold voltage in the Schmitt trigger. In this case, the noise carried in the signal after gain processing is used to drive the output of the Schmitt trigger to switch between high and low levels, forming a random resonance effect, thereby greatly enhancing the signal-to-noise ratio of the random resonance signal.

[0048] It should be noted that, in order to maximize the signal-to-noise ratio through the random resonance effect, in some embodiments, the processor generates gain with the goal of ensuring that the ratio between the root mean square of the noise in the signal after gain processing and the difference in the threshold voltage of the Schmitt trigger is within a pre-calibrated optimal ratio range.

[0049] In the processor, a gain is generated based on the output signal of the analog-to-digital converter (ADC). This includes statistically analyzing the switching frequency of the ADC output signal. If the switching frequency is lower than the pre-calibrated optimal frequency range, the gain is increased; if the switching frequency is higher than the pre-calibrated optimal frequency range, the gain is decreased; and if the switching frequency is within the pre-calibrated optimal frequency range, the gain remains unchanged.

[0050] In the processor, a reference threshold voltage is generated based on the output signal of the analog-to-digital converter, including: calculating the mean value of the output signal of the analog-to-digital converter; if the mean value is greater than 0.5, decreasing the reference threshold voltage; if the mean value is less than 0.5, increasing the reference threshold voltage.

[0051] The above circuit is added between the amplification and filtering module and the analog-to-digital converter of the existing balanced detection circuit. This method creates a random resonance effect, thereby greatly enhancing the signal-to-noise ratio of the random resonance signal.

[0052] This application also relates to a balanced detection method, which includes the above-mentioned signal-to-noise ratio enhancement method, specifically by adding a signal-to-noise ratio enhancement method to the existing balanced detection method.

[0053] This application also relates to a balance detection circuit, see [link to relevant documentation] Figure 4 The circuit may include at least an amplification and filtering module, the aforementioned signal-to-noise ratio enhancement circuit, an analog-to-digital converter, and a processor; the output of the amplification and filtering module is connected to the gain module of the signal-to-noise ratio enhancement circuit, the input of the analog-to-digital converter is connected to the output of the Schmitt trigger of the signal-to-noise ratio enhancement circuit, and the output of the analog-to-digital converter is connected to the processor.

[0054] It should be noted that the amplification and filtering module uses two operational amplifier chips, OPA847IDBVR. The first-stage operational amplifier is used to build a transimpedance amplifier circuit, and the second-stage operational amplifier is used to build a second-order low-pass filter circuit.

[0055] The input of the amplification and filtering module is connected to a balanced photoelectric detection circuit, which is used to convert the laser beat frequency signal of the coherent wind lidar into an electrical signal. The balanced photoelectric detection circuit mainly includes a resistor R1, two photodiodes and a resistor R2 connected in series. V+ and V- are the bias voltages of the photodiodes, which are connected to resistors R1 and R2 respectively. Resistors R1 and R2 are used to limit the output current of the photodiodes; the typical value of resistors R1 and R2 can be 50 ohms.

[0056] The processor can use the FPGA chip XC7K325T-2FFG900I. In addition to controlling the gain and reference threshold voltage, the processor also performs the function of balanced detection. Of course, depending on the actual situation, the processor can further process the sampled data, such as for inverting wind speed and wind direction information.

[0057] Analog-to-digital converters can be implemented using flip-flops within an FPGA chip. Since the comparator in the Schmitt trigger quantizes the signal, in some embodiments, a standalone single-bit analog-to-digital converter chip can be replaced with a single-bit analog-to-digital converter that is equivalent to a flip-flop and comparator. This helps to reduce circuit cost and hardware complexity.

[0058] In coherent wind lidar, the above circuit achieves balanced detection with enhanced signal-to-noise ratio, increasing the detection range by 3-5 km compared to traditional balanced detection circuits.

[0059] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of this invention, and these improvements and modifications should also be considered within the scope of protection of this invention.

Claims

1. A signal-to-noise ratio enhancement method, characterized in that, include: Gain processing is performed on the wind measurement laser radio frequency signal at the current moment; Based on the reference threshold voltage, the signal-to-noise ratio enhanced random resonance signal obtained at the previous moment, and the preset fixed relationship, a new threshold voltage is generated. The signal after gain processing at the current moment is compared with the new threshold voltage to obtain the signal-to-noise ratio enhanced random resonance signal at the current moment. Among them, the random resonance signal is input to the analog-to-digital converter of the balanced detector, the wind measurement laser radio frequency signal is the output signal of the amplification and filtering module of the balanced detector; the new threshold voltage is either a high threshold voltage or a low threshold voltage. The gain and reference threshold voltage are adjusted based on the analog-to-digital converter output signal. The goal of adjusting the reference threshold voltage is to make the average threshold voltage as close as possible to the average signal after gain processing. The goal of adjusting the gain is to make the peak-to-peak value of the signal after gain processing greater than the threshold voltage difference. The average threshold voltage is the average of the high threshold voltage and the low threshold voltage, and the threshold voltage difference is the difference between the high threshold voltage and the low threshold voltage.

2. The method according to claim 1, characterized in that, The goal of adjusting the gain also includes ensuring that the ratio between the root mean square of the noise in the signal after gain processing and the difference between the threshold voltage and the gain is within the range of the pre-calibrated optimal ratio.

3. The method according to claim 1 or 2, characterized in that, Adjusting the gain based on the analog-to-digital converter output signal includes: The gain is increased if the switching frequency of the analog-to-digital converter output signal is lower than the pre-calibrated optimal frequency range, and decreased if the switching frequency is higher than the pre-calibrated optimal frequency range. If the switching frequency is within the pre-calibrated optimal frequency range, the gain remains unchanged.

4. The method according to claim 1, characterized in that, Adjusting the reference threshold voltage based on the analog-to-digital converter output signal includes: Calculate the mean value of the analog-to-digital converter output signal. If the mean value is greater than 0.5, decrease the reference threshold voltage; if the mean value is less than 0.5, increase the reference threshold voltage.

5. A signal-to-noise ratio enhancement circuit, characterized in that, Includes gain modules and Schmitt triggers; The input terminal of the gain module is connected to the amplification and filtering module of the balanced detector. The first and second input terminals of the Schmitt trigger are connected to the output terminal of the gain module and the processor of the balanced detector, respectively. The output terminal of the Schmitt trigger is connected to the analog-to-digital converter of the balanced detector. The gain module is used to perform gain processing on the wind measurement laser radio frequency signal at the current moment; A Schmitt trigger is used to generate a new threshold voltage based on a reference threshold voltage, the signal-to-noise ratio enhanced random resonance signal obtained at the previous time step, and a preset fixed relationship; the signal after gain processing at the current time step is compared with the new threshold voltage to obtain the signal-to-noise ratio enhanced random resonance signal at the current time step; wherein, the new threshold voltage is a high threshold voltage or a low threshold voltage. The processor generates a gain and a reference threshold voltage based on the output signal of the analog-to-digital converter, and sends the generated gain and reference threshold voltage to the gain module and the Schmitt trigger, respectively. The processor generates a reference threshold voltage with the objective of making the mean threshold voltage as close as possible to the mean signal after gain processing. The processor generates a gain with the objective of making the peak-to-peak value of the signal after gain processing greater than the threshold voltage difference. The mean threshold voltage is the average of the high threshold voltage and the low threshold voltage, and the threshold voltage difference is the difference between the high threshold voltage and the low threshold voltage.

6. The circuit according to claim 5, characterized in that, The processor's goal in generating gain also includes ensuring that the ratio between the root mean square of noise in the signal after gain processing and the difference in threshold voltage is within a pre-calibrated optimal ratio range.

7. The circuit according to claim 5 or 6, characterized in that, In the processor, gain is generated based on the output signal of the analog-to-digital converter, including: The gain is increased if the switching frequency of the analog-to-digital converter output signal is lower than the pre-calibrated optimal frequency range, and decreased if the switching frequency is higher than the pre-calibrated optimal frequency range. If the switching frequency is within the pre-calibrated optimal frequency range, the gain remains unchanged.

8. The circuit according to claim 5, characterized in that, In the processor, a reference threshold voltage is generated based on the analog-to-digital converter output signal, including: Calculate the mean value of the analog-to-digital converter output signal. If the mean value is greater than 0.5, decrease the reference threshold voltage; if the mean value is less than 0.5, increase the reference threshold voltage.

9. A method for detecting balance, characterized in that, Includes the method described in any one of claims 1 to 4.

10. A balance detection circuit, characterized in that, Includes the circuit described in any one of claims 5 to 8.

Images