CN103067107B - Device and detection method based on superhet type stochastic resonance detection system - Google Patents

Abstract

The invention discloses to a device and a detection method based on a superhet type stochastic resonance detection system, and relates to a stochastic resonance detection device and a method which aims at solving the following two problems: (1) an existing stochastic resonance system is not suitable for a frequency signal where f 0 is greater than 1; and (2) a signal with a 108 magnitude frequency cannot be detected through a stochastic resonance system. An external signal to be tested is connected with an input end of a mixer, an signal output signal of the mixer is connected with an signal input end of an amplifier, an signal output end of the amplifier is connected a signal input end of a filter, an signal output end of the filter is connected with an signal input end of an analog-digital (A/D) converter, a data output end of the A/D converter is connected with a data input end of a bistable state stochastic resonance system leading into integral compensation, and data output end of the bistable state stochastic resonance system leading into the integral compensation is connected with a data input end of an external data processor. The stochastic resonance detection device and the method can be widely used for stochastic resonance detection for signals with the frequency of 1 to 108.

Description

Based on device and the detection method of superhet type stochastic resonance detection system

Technical field

The present invention relates to the device and method that a kind of accidental resonance detects, belong to signal detection technique field.

Background technology

At present, signal detection technique has in a lot of fields applies, very widely as biomedicine, electrochemistry, scientific research etc.Although people are in research Detection of Weak Signals, develop the Theories and methods of a whole set of restraint speckle, as digital lock-in technique, Boxcar integrator and coherent detection technology etc., sizable effect is served in Testing of Feeble Signals, but while restraint speckle, measured signal is also suppressed or loses.In addition, these methods are carry out in the framework of linear system mostly, as arrowbandization and coherent detection technology, the averaging of time-domain signal, the stroke analysis etc. of discrete signal.Therefore need to consider following Railway Project: 1 system itself is the source producing noise.In whole information process chain, the system related to is more, and the probability producing additional noise is also more.2 systems itself do not exist signal to noise ratio increases function.On the contrary, along with the increase of input noise, system output signal-to-noise ratio can constantly decline.Therefore a kind of Weak Signal Detection Method that can extract useful signal from noise is found to be the problem demanding prompt solution that society faces.

The proposition of Stochastic Resonance Theory, new thinking has been started in the detection for small-signal.Stochastic Resonance Theory shows, when small-signal under strong noise background is by a non linear system, if non-linear, the signal of system reach certain with noise and mate, background noise can strengthen the output of small-signal, improve output signal-to-noise ratio, but the limitation of stochastic resonance system to frequency input signal makes high-frequency signal detect to be hindered, and is only applicable to low frequency (f ₀< < 1Hz), to 10 ⁸the signal of magnitude frequency also cannot be detected by stochastic resonance system.

Summary of the invention

The present invention is not suitable for f to solve (1) existing stochastic resonance system ₀the frequency signal of >1; (2) 10 ⁸the signal of magnitude frequency cannot carry out by stochastic resonance system the two problems that detects, thus provides a kind of based on superhet accidental resonance checkout gear and method.

Based on the device of superhet type stochastic resonance detection system, it comprises frequency mixer, amplifier, filter, A/D converter, stochastic resonance system and local oscillation signal processor;

Outside measured signal is connected with the measured signal input of described frequency mixer, and the carrier signal output of local oscillation signal processor is connected with the carrier signal input of frequency mixer; The mixed frequency signal output of described frequency mixer is connected with the mixed frequency signal input of amplifier, the mixed frequency signal output of amplifier is connected with the mixed frequency signal input of filter, the down-scaled signals output of filter is connected with the down-scaled signals input of A/D converter, the frequency reducing data output end of A/D converter is connected with the frequency reducing data input pin of the bistable-state random resonance system introducing integral compensation, and the peak modal data output of bistable-state random resonance system introducing integral compensation is connected with the peak modal data input of external data processor;

The bistable-state random resonance system of described introducing integral compensation comprises bistable-state random resonance system and penalty coefficient module, and the penalty coefficient output of described penalty coefficient module is connected with the penalty coefficient input of stochastic resonance system.

Based on the method for the device of superhet type stochastic resonance detection system, it comprises the steps:

Step one: by measured signal s (t)+n (t) input mixer;

Described measured signal s (t)+n (t) for be mixed with white Gaussian noise original signal; Original signal is s (t)=Acos (2 π f ₀t), white Gaussian noise is n (t); The scope of described measured signal frequency is 1Hz ~ 9 × 10 ⁸hz;

Step 2: the carrier signal v that local oscillation signal processor produces from local oscillation signal processor _ct the lower-frequency limit of () starts to extract carrier signal v _c(t), the frequency interval of extraction is Δ f, and Δ f is non-zero arbitrary value; Obtain the carrier signal v of several extraction _c(t);

By the carrier signal v of each extraction _c(t) and measured signal s (t)+n (t) described in step one input mixer simultaneously, the initial time of measured signal s (t)+n (t) input mixer is t=0;

The carrier signal v that described local oscillation signal processor produces _ct () is v _c(t)=cos (2 π f _cand carrier signal v t), _ct the frequency range of () is identical with the frequency range of measured signal s (t)+n (t);

Step 3; By several carrier signal v of the extraction described in step 2 _ct () carries out mixing with measured signal s (t)+n (t) respectively, obtain several mixed frequency signal; Described measured signal s (t)+n (t) carrying out mixing is measured signal s (t)+n (t) of 0.1 second time cycle;

Described each mixed frequency signal is [s (t)+n (t)] v _ct (), through optical mixing process, obtains down-scaled signals x respectively ₁(t), up-conversion signal x ₂(t) and noise ξ (t); Described down-scaled signals x ₁(t) be up-conversion signal x ₂(t) be $x_2\left(t\right)=\frac{1}{2}A\cos \left[2\mathrm{\&pi;}(f_0+f_c)t\right];$

Step 4: several mixed frequency signals [s (t)+n (the t)] v that step 3 is obtained _ct () be input amplifier respectively, obtain several mixed frequency signals [s (t)+n (the t)] v after amplifying _c(t);

Step 5: several mixed frequency signals [s (t)+n (the t)] v after the amplification that step 4 is obtained _ct () be input filter respectively, obtain several mixed frequency signal [s (t)+n (t)] v _cseveral down-scaled signals x in (t) ₁(t);

Step 6: by down-scaled signals x several described in step 5 ₁t () input a/d converter, by A/D converter just down-scaled signals x ₁t () is converted to several down-scaled signals x ₁the digital signal of (t);

Step 7: several down-scaled signals x that step 6 is obtained ₁t the bistable-state random resonance system of integral compensation is introduced in the digital signal input of (), introduce the bistable-state random resonance system (6) of integral compensation by several down-scaled signals x ₁t the peak spectral amplitude of the digital signal of () exports ppu to.

Adopt the device and method that the present invention is based on superhet type stochastic resonance detection system, its advantage comprises:

(1) bistable-state random resonance system is introduce the stochastic resonance system of integral compensation, makes f ₀the signal of >1 produces accidental resonance;

(2) by introducing super-heterodyne technique, before entering bistable-state random resonance system, mixing being carried out to input signal, making the down-scaled signals of generation meet accidental resonance condition, further increase the frequency range that bistable-state random resonance system can detect.

Accompanying drawing explanation

Fig. 1 is the apparatus structure schematic diagram based on superhet type stochastic resonance detection system;

Fig. 2 is the method flow diagram based on superhet type stochastic resonance detection system.

Embodiment

Embodiment one, composition graphs 1 illustrate this embodiment.Based on the device of superhet type stochastic resonance detection system, it comprises frequency mixer 1, amplifier 3, filter 4, A/D converter 5, stochastic resonance system 6 and local oscillation signal processor 2;

Outside measured signal is connected with the measured signal input of described frequency mixer 1, and the carrier signal output of local oscillation signal processor 2 is connected with the carrier signal input of frequency mixer 1; The mixed frequency signal output of described frequency mixer 1 is connected with the mixed frequency signal input of amplifier 3, the mixed frequency signal output of amplifier 3 is connected with the mixed frequency signal input of filter 4, the down-scaled signals output of filter 4 is connected with the down-scaled signals input of A/D converter 5, the frequency reducing data output end of A/D converter 5 is connected with the frequency reducing data input pin of the bistable-state random resonance system 6 introducing integral compensation, and the peak modal data output of bistable-state random resonance system 6 introducing integral compensation is connected with the peak modal data input of external data processor 7;

The bistable-state random resonance system 6 of described introducing integral compensation comprises bistable-state random resonance system 8 and penalty coefficient module 9, and the penalty coefficient output of described penalty coefficient module 9 is connected with the penalty coefficient input of stochastic resonance system 8.

The detection method of the device based on superhet type stochastic resonance detection system of embodiment two, employing embodiment one, it comprises the steps:

Step one: by measured signal s (t)+n (t) input mixer 1;

Described measured signal s (t)+n (t) for be mixed with white Gaussian noise original signal; Original signal is s (t)=Acos (2 π f ₀t), white Gaussian noise is n (t); The scope of described measured signal frequency is 1Hz ~ 9 × 10 ⁸hz;

Step 2: the carrier signal v that local oscillation signal processor 2 produces from local oscillation signal processor 2 _ct the lower-frequency limit of () starts to extract carrier signal v _c(t), the frequency interval of extraction is Δ f, and Δ f is non-zero arbitrary value; Obtain the carrier signal v of several extraction _c(t);

By the carrier signal v of each extraction _c(t) and measured signal s (t)+n (t) described in step one input mixer 1 simultaneously, the initial time of measured signal s (t)+n (t) input mixer 1 is t=0;

The carrier signal v that described local oscillation signal processor 2 produces _ct () is v _c(t)=cos (2 π f _cand carrier signal v t), _ct the frequency range of () is identical with the frequency range of measured signal s (t)+n (t);

Step 3; By several carrier signal v of the extraction described in step 2 _ct () carries out mixing with measured signal s (t)+n (t) respectively, obtain several mixed frequency signal; Described measured signal s (t)+n (t) carrying out mixing is measured signal s (t)+n (t) of 0.1 second time cycle;

Described each mixed frequency signal is [s (t)+n (t)] v _ct (), through optical mixing process, obtains down-scaled signals x respectively ₁(t), up-conversion signal x ₂(t) and noise ξ (t); Described down-scaled signals x ₁(t) be up-conversion signal x ₂(t) be $x_2\left(t\right)=\frac{1}{2}A\cos \left[2\mathrm{\&pi;}(f_0+f_c)t\right];$

The principle of described mixing is:

$x\left(t\right)=[s\left(t\right)+n\left(t\right)]v_c\left(t\right)$

$=\frac{1}{2}A\cos \left[2\mathrm{\&pi;}(f_0-f_c)t\right]+\frac{1}{2}A\cos \left[2\mathrm{\&pi;}(f_0+f_c)t\right]+n\left(t\right)\cos \left(2\mathrm{\&pi;}{f}_{c}t\right)$

Step 4: several mixed frequency signals [s (t)+n (the t)] v that step 3 is obtained _ct () be input amplifier 3 respectively, obtain several mixed frequency signals [s (t)+n (the t)] v after amplifying _c(t);

Step 5: several mixed frequency signals [s (t)+n (the t)] v after the amplification that step 4 is obtained _ct () be input filter 4 respectively, obtain several mixed frequency signal [s (t)+n (t)] v _cseveral down-scaled signals x in (t) ₁(t);

Step 6: by the several down-scaled signals x described in step 5 ₁t () input a/d converter 5, by A/D converter 5 by down-scaled signals x ₁t () is converted to several down-scaled signals x ₁the digital signal of (t);

Step 7: several down-scaled signals x that step 6 is obtained ₁t the bistable-state random resonance system 6 of integral compensation is introduced in the digital signal input of (), introduce the bistable-state random resonance system 6 of integral compensation by several down-scaled signals x ₁t the peak spectral amplitude of the digital signal of () exports ppu 7 to.

Embodiment three, this embodiment and embodiment two unlike the principle of the bistable-state random resonance system 6 introducing integral compensation described in step 7 is:

Stochastic resonance system describes Brownian Particles under overdamp and acts on by cyclical signal and noise the reciprocal transition phenomenon occurred in non linear system.Equation-Langevin the equation describing Particles Moving is:

$\frac{d_x}{d_y}=\mathrm{ax}-{\mathrm{bx}}^3+\frac{1}{2}A\cos \left[2\mathrm{\&pi;}(f_0-f_c)t\right]+\mathrm{\&xi;}\left(t\right)$

The bistable-state random resonance system (6) introducing integral compensation is improved stochastic resonance system for introducing integral compensation:

$\frac{d_x}{d_y}=\{\mathrm{ax}-{\mathrm{bx}}^3+\frac{1}{2}A\cos [2\mathrm{\&pi;}(f_0-f_c)t]+\mathrm{\&xi;}\left(t\right)\}\&times;\mathrm{Gain}$

Wherein, Gain is integral compensation coefficient, and Gain>=2 π (f ₀-f ₀).

Stochastic resonance system adds integral compensation, and become the bistable-state random resonance system 6 introducing integral compensation, the signal frequency that can process increases to 10 ²hz magnitude.The relation of Time And Frequency is become frequency and amplitude relation by FFT shift conversion by stochastic resonance system, and is exported to external data processor 7.The frequency that each mixed frequency signal x (t) is obtained by stochastic resonance system by external data processor 7 and amplitude relation fit to continuous peak spectral amplitude, and calculate the frequency that Amplitude maxima is corresponding, described frequency is the frequency f of original signal ₀, described frequency is the carrier signal v extracted _cthe frequency f of (t) _c.Due to known as measured signal s (t)+n (t) and carrier signal v _cwhen the frequency of () is identical t, stochastic resonance system has maximum spectrum peak, namely by composing frequency f corresponding to peak maximum _cit is exactly the frequency f of measured signal s (t)+n (t) ₀, complete testing process.

Adopt the device and method that the present invention is based on superhet type stochastic resonance detection system, bistable-state random resonance system is introduce the stochastic resonance system of integral compensation, makes f ₀the signal of >1 produces accidental resonance; By introducing super-heterodyne technique, before entering bistable-state random resonance system, mixing being carried out to input signal, making the down-scaled signals of generation meet accidental resonance condition, further increase the frequency range that bistable-state random resonance system can detect.

Claims (2)

1. based on the detection method of the device of superhet type stochastic resonance detection system, the described device based on superhet type stochastic resonance detection system, it comprises frequency mixer (1), amplifier (3), filter (4), A/D converter (5), the bistable-state random resonance system (6) introducing integral compensation and local oscillation signal processor (2);

Outside measured signal is connected with the measured signal input of described frequency mixer (1), and the carrier signal output of local oscillation signal processor (2) is connected with the carrier signal input of frequency mixer (1), the mixed frequency signal output of described frequency mixer (1) is connected with the mixed frequency signal input of amplifier (3), the mixed frequency signal output of amplifier (3) is connected with the mixed frequency signal input of filter (4), the down-scaled signals output of filter (4) is connected with the down-scaled signals input of A/D converter (5), the frequency reducing data output end of A/D converter (5) is connected with the frequency reducing data input pin of the bistable-state random resonance system (6) introducing integral compensation, the peak modal data output of bistable-state random resonance system (6) introducing integral compensation is connected with the peak modal data input of external data processor (7),

The bistable-state random resonance system (6) of described introducing integral compensation comprises bistable-state random resonance system (8) and penalty coefficient module (9), and the penalty coefficient output of described penalty coefficient module (9) is connected with the penalty coefficient input of stochastic resonance system (8);

Based on the detection method of the device of superhet type stochastic resonance detection system, it is characterized in that it comprises the steps:

Step one: by measured signal s (t)+n (t) input mixer (1);

Described measured signal s (t)+n (t) for be mixed with white Gaussian noise original signal; Original signal is s (t)=Acos (2 π f ₀t), white Gaussian noise is n (t); The scope of described measured signal frequency is 1Hz ~ 9 × 10 ⁸hz;

Step 2: the carrier signal v that local oscillation signal processor (2) produces from local oscillation signal processor (2) _ct the lower-frequency limit of () starts to extract carrier signal v _c(t), the frequency interval of extraction is △ f, and △ f is non-zero arbitrary value; Obtain the carrier signal v of several extraction _c(t);

By the carrier signal v of each extraction _c(t) and measured signal s (t)+n (t) described in step one input mixer (1) simultaneously, the initial time of measured signal s (t)+n (t) input mixer (1) is t=0;

The carrier signal v that described local oscillation signal processor (2) produces _ct () is v _c(t)=cos (2 π f _cand carrier signal v t), _ct the frequency range of () is identical with the frequency range of measured signal s (t)+n (t);

Step 3; By several carrier signal v of the extraction described in step 2 _ct () carries out mixing with measured signal s (t)+n (t) respectively, obtain several mixed frequency signal; Described measured signal s (t)+n (t) carrying out mixing is measured signal s (t)+n (t) of 0.1 second time cycle:

Described each mixed frequency signal is [s (t)+n (t)] v _ct (), through optical mixing process, obtains down-scaled signals x respectively ₁(t), up-conversion signal x ₂(t) and noise ξ (t); Described down-scaled signals x ₁(t) be up-conversion signal x ₂(t) be $x_2\left(t\right)=\frac{1}{2}A\cos \left[2\mathrm{\&pi;}(f_0+f_c)t\right];$

Step 4: several mixed frequency signals [s (t)+n (the t)] v that step 3 is obtained _ct () be input amplifier (3) respectively, obtain several mixed frequency signals [s (t)+n (the t)] v after amplifying _c(t);

Step 5: several mixed frequency signals [s (t)+n (the t)] v after the amplification that step 4 is obtained _ct () be input filter (4) respectively, obtain several mixed frequency signal [s (t)+n (t)] v _cseveral down-scaled signals x in (t) ₁(t);

Step 6: by the several down-scaled signals x described in step 5 ₁t () input a/d converter (5), by A/D converter (5) by down-scaled signals x ₁t () is converted to several down-scaled signals x ₁the digital signal of (t);

Step 7: several down-scaled signals x that step 6 is obtained ₁t the bistable-state random resonance system (6) of integral compensation is introduced in the digital signal input of (), introduce the bistable-state random resonance system (6) of integral compensation by several down-scaled signals x ₁t the peak spectral amplitude of the digital signal of () exports external data processor (7) to.

2. the detection method of the device based on superhet type stochastic resonance detection system according to claim 1, is characterized in that the principle of the bistable-state random resonance system (6) introducing integral compensation described in step 7 is:

Stochastic resonance system describes Brownian Particles under overdamp and acts on by cyclical signal and noise the reciprocal transition phenomenon occurred in non linear system; Equation-Langevin the equation describing Particles Moving is:

$\frac{d_x}{d_y}=\mathrm{ax}-{\mathrm{bx}}^3+\frac{1}{2}A\cos \left[2\mathrm{\&pi;}(f_0-f_c)t\right]+\mathrm{\&xi;}\left(t\right)$

The bistable-state random resonance system (6) introducing integral compensation is improved stochastic resonance system for introducing integral compensation:

$\frac{d_x}{d_y}=\{\mathrm{ax}-{\mathrm{bx}}^3+\frac{1}{2}A\cos [2\mathrm{\&pi;}(f_0-f_c)t]+\mathrm{\&xi;}\left(t\right)\}\&times;\mathrm{Gain}$

Wherein, Gain is integral compensation coefficient, and Gain>=2 π (f ₀-f ).